4 from __future__
import unicode_literals
39 import xml
.etree
.ElementTree
44 compat_HTMLParseError
,
50 compat_ctypes_WINFUNCTYPE
,
51 compat_etree_fromstring
,
54 compat_html_entities_html5
,
68 compat_urllib_parse_urlencode
,
69 compat_urllib_parse_urlparse
,
70 compat_urllib_parse_urlunparse
,
71 compat_urllib_parse_quote
,
72 compat_urllib_parse_quote_plus
,
73 compat_urllib_parse_unquote_plus
,
74 compat_urllib_request
,
85 def register_socks_protocols():
86 # "Register" SOCKS protocols
87 # In Python < 2.6.5, urlsplit() suffers from bug https://bugs.python.org/issue7904
88 # URLs with protocols not in urlparse.uses_netloc are not handled correctly
89 for scheme
in ('socks', 'socks4', 'socks4a', 'socks5'):
90 if scheme
not in compat_urlparse
.uses_netloc
:
91 compat_urlparse
.uses_netloc
.append(scheme
)
94 # This is not clearly defined otherwise
95 compiled_regex_type
= type(re
.compile(''))
98 def random_user_agent():
99 _USER_AGENT_TPL
= 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/%s Safari/537.36'
140 return _USER_AGENT_TPL
% random
.choice(_CHROME_VERSIONS
)
144 'User-Agent': random_user_agent(),
145 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
146 'Accept-Encoding': 'gzip, deflate',
147 'Accept-Language': 'en-us,en;q=0.5',
148 'Sec-Fetch-Mode': 'navigate',
153 'Safari': 'Mozilla/5.0 (X11; Linux x86_64; rv:10.0) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27',
157 NO_DEFAULT
= object()
159 ENGLISH_MONTH_NAMES
= [
160 'January', 'February', 'March', 'April', 'May', 'June',
161 'July', 'August', 'September', 'October', 'November', 'December']
164 'en': ENGLISH_MONTH_NAMES
,
166 'janvier', 'février', 'mars', 'avril', 'mai', 'juin',
167 'juillet', 'août', 'septembre', 'octobre', 'novembre', 'décembre'],
171 'mp4', 'm4a', 'm4p', 'm4b', 'm4r', 'm4v', 'aac',
172 'flv', 'f4v', 'f4a', 'f4b',
173 'webm', 'ogg', 'ogv', 'oga', 'ogx', 'spx', 'opus',
174 'mkv', 'mka', 'mk3d',
183 'f4f', 'f4m', 'm3u8', 'smil')
185 # needed for sanitizing filenames in restricted mode
186 ACCENT_CHARS
= dict(zip('ÂÃÄÀÁÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖŐØŒÙÚÛÜŰÝÞßàáâãäåæçèéêëìíîïðñòóôõöőøœùúûüűýþÿ',
187 itertools
.chain('AAAAAA', ['AE'], 'CEEEEIIIIDNOOOOOOO', ['OE'], 'UUUUUY', ['TH', 'ss'],
188 'aaaaaa', ['ae'], 'ceeeeiiiionooooooo', ['oe'], 'uuuuuy', ['th'], 'y')))
218 '%Y-%m-%d %H:%M:%S.%f',
219 '%Y-%m-%d %H:%M:%S:%f',
222 '%Y-%m-%dT%H:%M:%SZ',
223 '%Y-%m-%dT%H:%M:%S.%fZ',
224 '%Y-%m-%dT%H:%M:%S.%f0Z',
226 '%Y-%m-%dT%H:%M:%S.%f',
229 '%b %d %Y at %H:%M:%S',
231 '%B %d %Y at %H:%M:%S',
235 DATE_FORMATS_DAY_FIRST
= list(DATE_FORMATS
)
236 DATE_FORMATS_DAY_FIRST
.extend([
245 DATE_FORMATS_MONTH_FIRST
= list(DATE_FORMATS
)
246 DATE_FORMATS_MONTH_FIRST
.extend([
254 PACKED_CODES_RE
= r
"}\('(.+)',(\d+),(\d+),'([^']+)'\.split\('\|'\)"
255 JSON_LD_RE
= r
'(?is)<script[^>]+type=(["\']?
)application
/ld\
+json\
1[^
>]*>(?P
<json_ld
>.+?
)</script
>'
258 def preferredencoding():
259 """Get preferred encoding.
261 Returns the best encoding scheme for the system, based on
262 locale.getpreferredencoding() and some further tweaks.
265 pref = locale.getpreferredencoding()
273 def write_json_file(obj, fn):
274 """ Encode obj as JSON and write it to fn, atomically if possible """
276 fn = encodeFilename(fn)
277 if sys.version_info < (3, 0) and sys.platform != 'win32
':
278 encoding = get_filesystem_encoding()
279 # os.path.basename returns a bytes object, but NamedTemporaryFile
280 # will fail if the filename contains non ascii characters unless we
281 # use a unicode object
282 path_basename = lambda f: os.path.basename(fn).decode(encoding)
283 # the same for os.path.dirname
284 path_dirname = lambda f: os.path.dirname(fn).decode(encoding)
286 path_basename = os.path.basename
287 path_dirname = os.path.dirname
291 'prefix
': path_basename(fn) + '.',
292 'dir': path_dirname(fn),
296 # In Python 2.x, json.dump expects a bytestream.
297 # In Python 3.x, it writes to a character stream
298 if sys.version_info < (3, 0):
306 tf = tempfile.NamedTemporaryFile(**compat_kwargs(args))
310 json.dump(obj, tf, ensure_ascii=False)
311 if sys.platform == 'win32
':
312 # Need to remove existing file on Windows, else os.rename raises
313 # WindowsError or FileExistsError.
321 os.chmod(tf.name, 0o666 & ~mask)
324 os.rename(tf.name, fn)
333 if sys.version_info >= (2, 7):
334 def find_xpath_attr(node, xpath, key, val=None):
335 """ Find the xpath xpath[@key=val] """
336 assert re.match(r'^
[a
-zA
-Z_
-]+$
', key)
337 expr = xpath + ('[@%s]' % key if val is None else "[@%s='%s']" % (key, val))
338 return node.find(expr)
340 def find_xpath_attr(node, xpath, key, val=None):
341 for f in node.findall(compat_xpath(xpath)):
342 if key not in f.attrib:
344 if val is None or f.attrib.get(key) == val:
348 # On python2.6 the xml.etree.ElementTree.Element methods don't support
349 # the namespace parameter
352 def xpath_with_ns(path
, ns_map
):
353 components
= [c
.split(':') for c
in path
.split('/')]
357 replaced
.append(c
[0])
360 replaced
.append('{%s}%s' % (ns_map
[ns
], tag
))
361 return '/'.join(replaced
)
364 def xpath_element(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
365 def _find_xpath(xpath
):
366 return node
.find(compat_xpath(xpath
))
368 if isinstance(xpath
, (str, compat_str
)):
369 n
= _find_xpath(xpath
)
377 if default
is not NO_DEFAULT
:
380 name
= xpath
if name
is None else name
381 raise ExtractorError('Could not find XML element %s' % name
)
387 def xpath_text(node
, xpath
, name
=None, fatal
=False, default
=NO_DEFAULT
):
388 n
= xpath_element(node
, xpath
, name
, fatal
=fatal
, default
=default
)
389 if n
is None or n
== default
:
392 if default
is not NO_DEFAULT
:
395 name
= xpath
if name
is None else name
396 raise ExtractorError('Could not find XML element\'s text %s' % name
)
402 def xpath_attr(node
, xpath
, key
, name
=None, fatal
=False, default
=NO_DEFAULT
):
403 n
= find_xpath_attr(node
, xpath
, key
)
405 if default
is not NO_DEFAULT
:
408 name
= '%s[@%s]' % (xpath
, key
) if name
is None else name
409 raise ExtractorError('Could not find XML attribute %s' % name
)
415 def get_element_by_id(id, html
):
416 """Return the content of the tag with the specified ID in the passed HTML document"""
417 return get_element_by_attribute('id', id, html
)
420 def get_element_html_by_id(id, html
):
421 """Return the html of the tag with the specified ID in the passed HTML document"""
422 return get_element_html_by_attribute('id', id, html
)
425 def get_element_by_class(class_name
, html
):
426 """Return the content of the first tag with the specified class in the passed HTML document"""
427 retval
= get_elements_by_class(class_name
, html
)
428 return retval
[0] if retval
else None
431 def get_element_html_by_class(class_name
, html
):
432 """Return the html of the first tag with the specified class in the passed HTML document"""
433 retval
= get_elements_html_by_class(class_name
, html
)
434 return retval
[0] if retval
else None
437 def get_element_by_attribute(attribute
, value
, html
, escape_value
=True):
438 retval
= get_elements_by_attribute(attribute
, value
, html
, escape_value
)
439 return retval
[0] if retval
else None
442 def get_element_html_by_attribute(attribute
, value
, html
, escape_value
=True):
443 retval
= get_elements_html_by_attribute(attribute
, value
, html
, escape_value
)
444 return retval
[0] if retval
else None
447 def get_elements_by_class(class_name
, html
):
448 """Return the content of all tags with the specified class in the passed HTML document as a list"""
449 return get_elements_by_attribute(
450 'class', r
'[^\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
451 html, escape_value=False)
454 def get_elements_html_by_class(class_name, html):
455 """Return the html of all tags with the specified class in the passed HTML document as a list"""
456 return get_elements_html_by_attribute(
457 'class', r'[^
\'"]*\b%s\b[^\'"]*' % re.escape(class_name),
458 html, escape_value=False)
461 def get_elements_by_attribute(*args, **kwargs):
462 """Return the content of the tag with the specified attribute in the passed HTML document"""
463 return [content for content, _ in get_elements_text_and_html_by_attribute(*args, **kwargs)]
466 def get_elements_html_by_attribute(*args, **kwargs):
467 """Return the html of the tag with the specified attribute in the passed HTML document"""
468 return [whole for _, whole in get_elements_text_and_html_by_attribute(*args, **kwargs)]
471 def get_elements_text_and_html_by_attribute(attribute, value, html, escape_value=True):
473 Return the text (content) and the html (whole) of the tag with the specified
474 attribute in the passed HTML document
477 value_quote_optional = '' if re.match(r'''[\s"'`
=<>]''', value) else '?'
479 value = re.escape(value) if escape_value else value
481 partial_element_re = r'''(?x
)
482 <(?P
<tag
>[a
-zA
-Z0
-9:._-]+)
483 (?
:\
s(?
:[^
>"']|"[^
"]*"|
'[^']*')*)?
484 \s%(attribute)s\s*=\s*(?P<_q>['"]%(vqo)s)(?-x:%(value)s)(?P=_q)
485 ''' % {'attribute': re.escape(attribute), 'value': value, 'vqo': value_quote_optional}
487 for m in re.finditer(partial_element_re, html):
488 content, whole = get_element_text_and_html_by_tag(m.group('tag'), html[m.start():])
491 unescapeHTML(re.sub(r'^(?P<q>["\'])(?P
<content
>.*)(?P
=q
)$
', r'\g
<content
>', content, flags=re.DOTALL)),
496 class HTMLBreakOnClosingTagParser(compat_HTMLParser):
498 HTML parser which raises HTMLBreakOnClosingTagException upon reaching the
499 closing tag for the first opening tag it has encountered, and can be used
503 class HTMLBreakOnClosingTagException(Exception):
507 self.tagstack = collections.deque()
508 compat_HTMLParser.__init__(self)
513 def __exit__(self, *_):
517 # handle_endtag does not return upon raising HTMLBreakOnClosingTagException,
518 # so data remains buffered; we no longer have any interest in it, thus
519 # override this method to discard it
522 def handle_starttag(self, tag, _):
523 self.tagstack.append(tag)
525 def handle_endtag(self, tag):
526 if not self.tagstack:
527 raise compat_HTMLParseError('no tags
in the stack
')
529 inner_tag = self.tagstack.pop()
533 raise compat_HTMLParseError(f'matching opening tag
for closing {tag} tag
not found
')
534 if not self.tagstack:
535 raise self.HTMLBreakOnClosingTagException()
538 def get_element_text_and_html_by_tag(tag, html):
540 For the first element with the specified tag in the passed HTML document
541 return its' content (text
) and the whole
element (html
)
543 def find_or_raise(haystack, needle, exc):
545 return haystack.index(needle)
548 closing_tag = f'</{tag}>'
549 whole_start = find_or_raise(
550 html, f'<{tag}', compat_HTMLParseError(f'opening {tag} tag not found'))
551 content_start = find_or_raise(
552 html[whole_start:], '>', compat_HTMLParseError(f'malformed opening {tag} tag'))
553 content_start += whole_start + 1
554 with HTMLBreakOnClosingTagParser() as parser:
555 parser.feed(html[whole_start:content_start])
556 if not parser.tagstack or parser.tagstack[0] != tag:
557 raise compat_HTMLParseError(f'parser did not match opening {tag} tag')
558 offset = content_start
559 while offset < len(html):
560 next_closing_tag_start = find_or_raise(
561 html[offset:], closing_tag,
562 compat_HTMLParseError(f'closing {tag} tag not found'))
563 next_closing_tag_end = next_closing_tag_start + len(closing_tag)
565 parser.feed(html[offset:offset + next_closing_tag_end])
566 offset += next_closing_tag_end
567 except HTMLBreakOnClosingTagParser.HTMLBreakOnClosingTagException:
568 return html[content_start:offset + next_closing_tag_start], \
569 html[whole_start:offset + next_closing_tag_end]
570 raise compat_HTMLParseError('unexpected end of html')
573 class HTMLAttributeParser(compat_HTMLParser):
574 """Trivial HTML parser to gather the attributes
for a single element
"""
578 compat_HTMLParser.__init__(self)
580 def handle_starttag(self, tag, attrs):
581 self.attrs = dict(attrs)
584 class HTMLListAttrsParser(compat_HTMLParser):
585 """HTML parser to gather the attributes
for the elements of a
list"""
588 compat_HTMLParser.__init__(self)
592 def handle_starttag(self, tag, attrs):
593 if tag == 'li' and self._level == 0:
594 self.items.append(dict(attrs))
597 def handle_endtag(self, tag):
601 def extract_attributes(html_element):
602 """Given a string
for an HTML element such
as
604 a
="foo" B
="bar" c
="&98;az" d
=boz
605 empty
= noval entity
="&"
608 Decode
and return a dictionary of attributes
.
610 'a': 'foo', 'b': 'bar', c
: 'baz', d
: 'boz',
611 'empty': '', 'noval': None, 'entity': '&',
612 'sq': '"', 'dq': '\''
614 NB HTMLParser
is stricter
in Python
2.6 & 3.2 than
in later versions
,
615 but the cases
in the unit test will work
for all of
2.6, 2.7, 3.2-3.5.
617 parser = HTMLAttributeParser()
619 parser.feed(html_element)
621 # Older Python may throw HTMLParseError in case of malformed HTML
622 except compat_HTMLParseError:
627 def parse_list(webpage):
628 """Given a string
for an series of HTML
<li
> elements
,
629 return a dictionary of their attributes
"""
630 parser = HTMLListAttrsParser()
636 def clean_html(html):
637 """Clean an HTML snippet into a readable string
"""
639 if html is None: # Convenience for sanitizing descriptions etc.
642 html = re.sub(r'\s+', ' ', html)
643 html = re.sub(r'(?u)\s?<\s?br\s?/?\s?>\s?', '\n', html)
644 html = re.sub(r'(?u)<\s?/\s?p\s?>\s?<\s?p[^>]*>', '\n', html)
646 html = re.sub('<.*?>', '', html)
647 # Replace html entities
648 html = unescapeHTML(html)
652 def sanitize_open(filename, open_mode):
653 """Try to
open the given filename
, and slightly tweak it
if this fails
.
655 Attempts to
open the given filename
. If this fails
, it tries to change
656 the filename slightly
, step by step
, until it
's either able to open it
657 or it fails and raises a final exception, like the standard open()
660 It returns the tuple (stream, definitive_file_name).
664 if sys.platform == 'win32
':
666 msvcrt.setmode(sys.stdout.fileno(), os.O_BINARY)
667 return (sys.stdout.buffer if hasattr(sys.stdout, 'buffer') else sys.stdout, filename)
668 stream = open(encodeFilename(filename), open_mode)
669 return (stream, filename)
670 except (IOError, OSError) as err:
671 if err.errno in (errno.EACCES,):
674 # In case of error, try to remove win32 forbidden chars
675 alt_filename = sanitize_path(filename)
676 if alt_filename == filename:
679 # An exception here should be caught in the caller
680 stream = open(encodeFilename(alt_filename), open_mode)
681 return (stream, alt_filename)
684 def timeconvert(timestr):
685 """Convert RFC 2822 defined time string into system timestamp"""
687 timetuple = email.utils.parsedate_tz(timestr)
688 if timetuple is not None:
689 timestamp = email.utils.mktime_tz(timetuple)
693 def sanitize_filename(s, restricted=False, is_id=False):
694 """Sanitizes a string so it could be used as part of a filename.
695 If restricted is set, use a stricter subset of allowed characters.
696 Set is_id if this is not an arbitrary string, but an ID that should be kept
699 def replace_insane(char):
700 if restricted and char in ACCENT_CHARS:
701 return ACCENT_CHARS[char]
702 elif not restricted and char == '\n':
704 elif char == '?
' or ord(char) < 32 or ord(char) == 127:
707 return '' if restricted else '\''
709 return '_-' if restricted else ' -'
710 elif char in '\\/|*<>':
712 if restricted and (char in '!&\'()[]{}$;`^,#' or char.isspace()):
714 if restricted and ord(char) > 127:
721 s = re.sub(r'[0-9]+(?::[0-9]+)+', lambda m: m.group(0).replace(':', '_'), s)
722 result = ''.join(map(replace_insane, s))
724 while '__' in result:
725 result = result.replace('__', '_')
726 result = result.strip('_')
727 # Common case of "Foreign band name
- English song title
"
728 if restricted and result.startswith('-_'):
730 if result.startswith('-'):
731 result = '_' + result[len('-'):]
732 result = result.lstrip('.')
738 def sanitize_path(s, force=False):
739 """Sanitizes and normalizes path on Windows"""
740 if sys.platform == 'win32':
742 drive_or_unc, _ = os.path.splitdrive(s)
743 if sys.version_info < (2, 7) and not drive_or_unc:
744 drive_or_unc, _ = os.path.splitunc(s)
750 norm_path = os.path.normpath(remove_start(s, drive_or_unc)).split(os.path.sep)
754 path_part if path_part in ['.', '..'] else re.sub(r'(?:[/<>:"\|
\\?\
*]|
[\s
.]$
)', '#', path_part)
755 for path_part
in norm_path
]
757 sanitized_path
.insert(0, drive_or_unc
+ os
.path
.sep
)
758 elif force
and s
[0] == os
.path
.sep
:
759 sanitized_path
.insert(0, os
.path
.sep
)
760 return os
.path
.join(*sanitized_path
)
763 def sanitize_url(url
):
764 # Prepend protocol-less URLs with `http:` scheme in order to mitigate
765 # the number of unwanted failures due to missing protocol
766 if url
.startswith('//'):
767 return 'http:%s' % url
768 # Fix some common typos seen so far
770 # https://github.com/ytdl-org/youtube-dl/issues/15649
771 (r
'^httpss://', r
'https://'),
772 # https://bx1.be/lives/direct-tv/
773 (r
'^rmtp([es]?)://', r
'rtmp\1://'),
775 for mistake
, fixup
in COMMON_TYPOS
:
776 if re
.match(mistake
, url
):
777 return re
.sub(mistake
, fixup
, url
)
781 def extract_basic_auth(url
):
782 parts
= compat_urlparse
.urlsplit(url
)
783 if parts
.username
is None:
785 url
= compat_urlparse
.urlunsplit(parts
._replace
(netloc
=(
786 parts
.hostname
if parts
.port
is None
787 else '%s:%d' % (parts
.hostname
, parts
.port
))))
788 auth_payload
= base64
.b64encode(
789 ('%s:%s' % (parts
.username
, parts
.password
or '')).encode('utf-8'))
790 return url
, 'Basic ' + auth_payload
.decode('utf-8')
793 def sanitized_Request(url
, *args
, **kwargs
):
794 url
, auth_header
= extract_basic_auth(escape_url(sanitize_url(url
)))
795 if auth_header
is not None:
796 headers
= args
[1] if len(args
) >= 2 else kwargs
.setdefault('headers', {})
797 headers
['Authorization'] = auth_header
798 return compat_urllib_request
.Request(url
, *args
, **kwargs
)
802 """Expand shell variables and ~"""
803 return os
.path
.expandvars(compat_expanduser(s
))
806 def orderedSet(iterable
):
807 """ Remove all duplicates from the input iterable """
815 def _htmlentity_transform(entity_with_semicolon
):
816 """Transforms an HTML entity to a character."""
817 entity
= entity_with_semicolon
[:-1]
819 # Known non-numeric HTML entity
820 if entity
in compat_html_entities
.name2codepoint
:
821 return compat_chr(compat_html_entities
.name2codepoint
[entity
])
823 # TODO: HTML5 allows entities without a semicolon. For example,
824 # 'Éric' should be decoded as 'Éric'.
825 if entity_with_semicolon
in compat_html_entities_html5
:
826 return compat_html_entities_html5
[entity_with_semicolon
]
828 mobj
= re
.match(r
'#(x[0-9a-fA-F]+|[0-9]+)', entity
)
830 numstr
= mobj
.group(1)
831 if numstr
.startswith('x'):
833 numstr
= '0%s' % numstr
836 # See https://github.com/ytdl-org/youtube-dl/issues/7518
838 return compat_chr(int(numstr
, base
))
842 # Unknown entity in name, return its literal representation
843 return '&%s;' % entity
849 assert type(s
) == compat_str
852 r
'&([^&;]+;)', lambda m
: _htmlentity_transform(m
.group(1)), s
)
855 def escapeHTML(text
):
858 .replace('&', '&')
859 .replace('<', '<')
860 .replace('>', '>')
861 .replace('"', '"')
862 .replace("'", ''')
866 def process_communicate_or_kill(p
, *args
, **kwargs
):
868 return p
.communicate(*args
, **kwargs
)
869 except BaseException
: # Including KeyboardInterrupt
875 class Popen(subprocess
.Popen
):
876 if sys
.platform
== 'win32':
877 _startupinfo
= subprocess
.STARTUPINFO()
878 _startupinfo
.dwFlags |
= subprocess
.STARTF_USESHOWWINDOW
882 def __init__(self
, *args
, **kwargs
):
883 super(Popen
, self
).__init
__(*args
, **kwargs
, startupinfo
=self
._startupinfo
)
885 def communicate_or_kill(self
, *args
, **kwargs
):
886 return process_communicate_or_kill(self
, *args
, **kwargs
)
889 def get_subprocess_encoding():
890 if sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
891 # For subprocess calls, encode with locale encoding
892 # Refer to http://stackoverflow.com/a/9951851/35070
893 encoding
= preferredencoding()
895 encoding
= sys
.getfilesystemencoding()
901 def encodeFilename(s
, for_subprocess
=False):
903 @param s The name of the file
906 assert type(s
) == compat_str
908 # Python 3 has a Unicode API
909 if sys
.version_info
>= (3, 0):
912 # Pass '' directly to use Unicode APIs on Windows 2000 and up
913 # (Detecting Windows NT 4 is tricky because 'major >= 4' would
914 # match Windows 9x series as well. Besides, NT 4 is obsolete.)
915 if not for_subprocess
and sys
.platform
== 'win32' and sys
.getwindowsversion()[0] >= 5:
918 # Jython assumes filenames are Unicode strings though reported as Python 2.x compatible
919 if sys
.platform
.startswith('java'):
922 return s
.encode(get_subprocess_encoding(), 'ignore')
925 def decodeFilename(b
, for_subprocess
=False):
927 if sys
.version_info
>= (3, 0):
930 if not isinstance(b
, bytes):
933 return b
.decode(get_subprocess_encoding(), 'ignore')
936 def encodeArgument(s
):
937 if not isinstance(s
, compat_str
):
938 # Legacy code that uses byte strings
939 # Uncomment the following line after fixing all post processors
940 # assert False, 'Internal error: %r should be of type %r, is %r' % (s, compat_str, type(s))
941 s
= s
.decode('ascii')
942 return encodeFilename(s
, True)
945 def decodeArgument(b
):
946 return decodeFilename(b
, True)
949 def decodeOption(optval
):
952 if isinstance(optval
, bytes):
953 optval
= optval
.decode(preferredencoding())
955 assert isinstance(optval
, compat_str
)
959 _timetuple
= collections
.namedtuple('Time', ('hours', 'minutes', 'seconds', 'milliseconds'))
962 def timetuple_from_msec(msec
):
963 secs
, msec
= divmod(msec
, 1000)
964 mins
, secs
= divmod(secs
, 60)
965 hrs
, mins
= divmod(mins
, 60)
966 return _timetuple(hrs
, mins
, secs
, msec
)
969 def formatSeconds(secs
, delim
=':', msec
=False):
970 time
= timetuple_from_msec(secs
* 1000)
972 ret
= '%d%s%02d%s%02d' % (time
.hours
, delim
, time
.minutes
, delim
, time
.seconds
)
974 ret
= '%d%s%02d' % (time
.minutes
, delim
, time
.seconds
)
976 ret
= '%d' % time
.seconds
977 return '%s.%03d' % (ret
, time
.milliseconds
) if msec
else ret
980 def _ssl_load_windows_store_certs(ssl_context
, storename
):
981 # Code adapted from _load_windows_store_certs in https://github.com/python/cpython/blob/main/Lib/ssl.py
983 certs
= [cert
for cert
, encoding
, trust
in ssl
.enum_certificates(storename
)
984 if encoding
== 'x509_asn' and (
985 trust
is True or ssl
.Purpose
.SERVER_AUTH
.oid
in trust
)]
986 except PermissionError
:
990 ssl_context
.load_verify_locations(cadata
=cert
)
995 def make_HTTPS_handler(params
, **kwargs
):
996 opts_check_certificate
= not params
.get('nocheckcertificate')
997 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
998 context
.check_hostname
= opts_check_certificate
999 if params
.get('legacyserverconnect'):
1000 context
.options |
= 4 # SSL_OP_LEGACY_SERVER_CONNECT
1001 context
.verify_mode
= ssl
.CERT_REQUIRED
if opts_check_certificate
else ssl
.CERT_NONE
1002 if opts_check_certificate
:
1004 context
.load_default_certs()
1005 # Work around the issue in load_default_certs when there are bad certificates. See:
1006 # https://github.com/yt-dlp/yt-dlp/issues/1060,
1007 # https://bugs.python.org/issue35665, https://bugs.python.org/issue45312
1008 except ssl
.SSLError
:
1009 # enum_certificates is not present in mingw python. See https://github.com/yt-dlp/yt-dlp/issues/1151
1010 if sys
.platform
== 'win32' and hasattr(ssl
, 'enum_certificates'):
1011 # Create a new context to discard any certificates that were already loaded
1012 context
= ssl
.SSLContext(ssl
.PROTOCOL_TLS_CLIENT
)
1013 context
.check_hostname
, context
.verify_mode
= True, ssl
.CERT_REQUIRED
1014 for storename
in ('CA', 'ROOT'):
1015 _ssl_load_windows_store_certs(context
, storename
)
1016 context
.set_default_verify_paths()
1017 return YoutubeDLHTTPSHandler(params
, context
=context
, **kwargs
)
1020 def bug_reports_message(before
=';'):
1021 if ytdl_is_updateable():
1022 update_cmd
= 'type yt-dlp -U to update'
1024 update_cmd
= 'see https://github.com/yt-dlp/yt-dlp on how to update'
1025 msg
= 'please report this issue on https://github.com/yt-dlp/yt-dlp .'
1026 msg
+= ' Make sure you are using the latest version; %s.' % update_cmd
1027 msg
+= ' Be sure to call yt-dlp with the --verbose flag and include its complete output.'
1029 before
= before
.rstrip()
1030 if not before
or before
.endswith(('.', '!', '?')):
1031 msg
= msg
[0].title() + msg
[1:]
1033 return (before
+ ' ' if before
else '') + msg
1036 class YoutubeDLError(Exception):
1037 """Base exception for YoutubeDL errors."""
1040 def __init__(self
, msg
=None):
1043 elif self
.msg
is None:
1044 self
.msg
= type(self
).__name
__
1045 super().__init
__(self
.msg
)
1048 network_exceptions
= [compat_urllib_error
.URLError
, compat_http_client
.HTTPException
, socket
.error
]
1049 if hasattr(ssl
, 'CertificateError'):
1050 network_exceptions
.append(ssl
.CertificateError
)
1051 network_exceptions
= tuple(network_exceptions
)
1054 class ExtractorError(YoutubeDLError
):
1055 """Error during info extraction."""
1057 def __init__(self
, msg
, tb
=None, expected
=False, cause
=None, video_id
=None, ie
=None):
1058 """ tb, if given, is the original traceback (so that it can be printed out).
1059 If expected is set, this is a normal error message and most likely not a bug in yt-dlp.
1061 if sys
.exc_info()[0] in network_exceptions
:
1066 self
.expected
= expected
1068 self
.video_id
= video_id
1070 self
.exc_info
= sys
.exc_info() # preserve original exception
1072 super(ExtractorError
, self
).__init
__(''.join((
1073 format_field(ie
, template
='[%s] '),
1074 format_field(video_id
, template
='%s: '),
1076 format_field(cause
, template
=' (caused by %r)'),
1077 '' if expected
else bug_reports_message())))
1079 def format_traceback(self
):
1080 if self
.traceback
is None:
1082 return ''.join(traceback
.format_tb(self
.traceback
))
1085 class UnsupportedError(ExtractorError
):
1086 def __init__(self
, url
):
1087 super(UnsupportedError
, self
).__init
__(
1088 'Unsupported URL: %s' % url
, expected
=True)
1092 class RegexNotFoundError(ExtractorError
):
1093 """Error when a regex didn't match"""
1097 class GeoRestrictedError(ExtractorError
):
1098 """Geographic restriction Error exception.
1100 This exception may be thrown when a video is not available from your
1101 geographic location due to geographic restrictions imposed by a website.
1104 def __init__(self
, msg
, countries
=None, **kwargs
):
1105 kwargs
['expected'] = True
1106 super(GeoRestrictedError
, self
).__init
__(msg
, **kwargs
)
1107 self
.countries
= countries
1110 class DownloadError(YoutubeDLError
):
1111 """Download Error exception.
1113 This exception may be thrown by FileDownloader objects if they are not
1114 configured to continue on errors. They will contain the appropriate
1118 def __init__(self
, msg
, exc_info
=None):
1119 """ exc_info, if given, is the original exception that caused the trouble (as returned by sys.exc_info()). """
1120 super(DownloadError
, self
).__init
__(msg
)
1121 self
.exc_info
= exc_info
1124 class EntryNotInPlaylist(YoutubeDLError
):
1125 """Entry not in playlist exception.
1127 This exception will be thrown by YoutubeDL when a requested entry
1128 is not found in the playlist info_dict
1130 msg
= 'Entry not found in info'
1133 class SameFileError(YoutubeDLError
):
1134 """Same File exception.
1136 This exception will be thrown by FileDownloader objects if they detect
1137 multiple files would have to be downloaded to the same file on disk.
1139 msg
= 'Fixed output name but more than one file to download'
1141 def __init__(self
, filename
=None):
1142 if filename
is not None:
1143 self
.msg
+= f
': {filename}'
1144 super().__init
__(self
.msg
)
1147 class PostProcessingError(YoutubeDLError
):
1148 """Post Processing exception.
1150 This exception may be raised by PostProcessor's .run() method to
1151 indicate an error in the postprocessing task.
1155 class DownloadCancelled(YoutubeDLError
):
1156 """ Exception raised when the download queue should be interrupted """
1157 msg
= 'The download was cancelled'
1160 class ExistingVideoReached(DownloadCancelled
):
1161 """ --break-on-existing triggered """
1162 msg
= 'Encountered a video that is already in the archive, stopping due to --break-on-existing'
1165 class RejectedVideoReached(DownloadCancelled
):
1166 """ --break-on-reject triggered """
1167 msg
= 'Encountered a video that did not match filter, stopping due to --break-on-reject'
1170 class MaxDownloadsReached(DownloadCancelled
):
1171 """ --max-downloads limit has been reached. """
1172 msg
= 'Maximum number of downloads reached, stopping due to --max-downloads'
1175 class ReExtractInfo(YoutubeDLError
):
1176 """ Video info needs to be re-extracted. """
1178 def __init__(self
, msg
, expected
=False):
1179 super().__init
__(msg
)
1180 self
.expected
= expected
1183 class ThrottledDownload(ReExtractInfo
):
1184 """ Download speed below --throttled-rate. """
1185 msg
= 'The download speed is below throttle limit'
1188 super().__init
__(self
.msg
, expected
=False)
1191 class UnavailableVideoError(YoutubeDLError
):
1192 """Unavailable Format exception.
1194 This exception will be thrown when a video is requested
1195 in a format that is not available for that video.
1197 msg
= 'Unable to download video'
1199 def __init__(self
, err
=None):
1201 self
.msg
+= f
': {err}'
1202 super().__init
__(self
.msg
)
1205 class ContentTooShortError(YoutubeDLError
):
1206 """Content Too Short exception.
1208 This exception may be raised by FileDownloader objects when a file they
1209 download is too small for what the server announced first, indicating
1210 the connection was probably interrupted.
1213 def __init__(self
, downloaded
, expected
):
1214 super(ContentTooShortError
, self
).__init
__(
1215 'Downloaded {0} bytes, expected {1} bytes'.format(downloaded
, expected
)
1218 self
.downloaded
= downloaded
1219 self
.expected
= expected
1222 class XAttrMetadataError(YoutubeDLError
):
1223 def __init__(self
, code
=None, msg
='Unknown error'):
1224 super(XAttrMetadataError
, self
).__init
__(msg
)
1228 # Parsing code and msg
1229 if (self
.code
in (errno
.ENOSPC
, errno
.EDQUOT
)
1230 or 'No space left' in self
.msg
or 'Disk quota exceeded' in self
.msg
):
1231 self
.reason
= 'NO_SPACE'
1232 elif self
.code
== errno
.E2BIG
or 'Argument list too long' in self
.msg
:
1233 self
.reason
= 'VALUE_TOO_LONG'
1235 self
.reason
= 'NOT_SUPPORTED'
1238 class XAttrUnavailableError(YoutubeDLError
):
1242 def _create_http_connection(ydl_handler
, http_class
, is_https
, *args
, **kwargs
):
1243 # Working around python 2 bug (see http://bugs.python.org/issue17849) by limiting
1244 # expected HTTP responses to meet HTTP/1.0 or later (see also
1245 # https://github.com/ytdl-org/youtube-dl/issues/6727)
1246 if sys
.version_info
< (3, 0):
1247 kwargs
['strict'] = True
1248 hc
= http_class(*args
, **compat_kwargs(kwargs
))
1249 source_address
= ydl_handler
._params
.get('source_address')
1251 if source_address
is not None:
1252 # This is to workaround _create_connection() from socket where it will try all
1253 # address data from getaddrinfo() including IPv6. This filters the result from
1254 # getaddrinfo() based on the source_address value.
1255 # This is based on the cpython socket.create_connection() function.
1256 # https://github.com/python/cpython/blob/master/Lib/socket.py#L691
1257 def _create_connection(address
, timeout
=socket
._GLOBAL
_DEFAULT
_TIMEOUT
, source_address
=None):
1258 host
, port
= address
1260 addrs
= socket
.getaddrinfo(host
, port
, 0, socket
.SOCK_STREAM
)
1261 af
= socket
.AF_INET
if '.' in source_address
[0] else socket
.AF_INET6
1262 ip_addrs
= [addr
for addr
in addrs
if addr
[0] == af
]
1263 if addrs
and not ip_addrs
:
1264 ip_version
= 'v4' if af
== socket
.AF_INET
else 'v6'
1266 "No remote IP%s addresses available for connect, can't use '%s' as source address"
1267 % (ip_version
, source_address
[0]))
1268 for res
in ip_addrs
:
1269 af
, socktype
, proto
, canonname
, sa
= res
1272 sock
= socket
.socket(af
, socktype
, proto
)
1273 if timeout
is not socket
._GLOBAL
_DEFAULT
_TIMEOUT
:
1274 sock
.settimeout(timeout
)
1275 sock
.bind(source_address
)
1277 err
= None # Explicitly break reference cycle
1279 except socket
.error
as _
:
1281 if sock
is not None:
1286 raise socket
.error('getaddrinfo returns an empty list')
1287 if hasattr(hc
, '_create_connection'):
1288 hc
._create
_connection
= _create_connection
1289 sa
= (source_address
, 0)
1290 if hasattr(hc
, 'source_address'): # Python 2.7+
1291 hc
.source_address
= sa
1293 def _hc_connect(self
, *args
, **kwargs
):
1294 sock
= _create_connection(
1295 (self
.host
, self
.port
), self
.timeout
, sa
)
1297 self
.sock
= ssl
.wrap_socket(
1298 sock
, self
.key_file
, self
.cert_file
,
1299 ssl_version
=ssl
.PROTOCOL_TLSv1
)
1302 hc
.connect
= functools
.partial(_hc_connect
, hc
)
1307 def handle_youtubedl_headers(headers
):
1308 filtered_headers
= headers
1310 if 'Youtubedl-no-compression' in filtered_headers
:
1311 filtered_headers
= dict((k
, v
) for k
, v
in filtered_headers
.items() if k
.lower() != 'accept-encoding')
1312 del filtered_headers
['Youtubedl-no-compression']
1314 return filtered_headers
1317 class YoutubeDLHandler(compat_urllib_request
.HTTPHandler
):
1318 """Handler for HTTP requests and responses.
1320 This class, when installed with an OpenerDirector, automatically adds
1321 the standard headers to every HTTP request and handles gzipped and
1322 deflated responses from web servers. If compression is to be avoided in
1323 a particular request, the original request in the program code only has
1324 to include the HTTP header "Youtubedl-no-compression", which will be
1325 removed before making the real request.
1327 Part of this code was copied from:
1329 http://techknack.net/python-urllib2-handlers/
1331 Andrew Rowls, the author of that code, agreed to release it to the
1335 def __init__(self
, params
, *args
, **kwargs
):
1336 compat_urllib_request
.HTTPHandler
.__init
__(self
, *args
, **kwargs
)
1337 self
._params
= params
1339 def http_open(self
, req
):
1340 conn_class
= compat_http_client
.HTTPConnection
1342 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1344 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1345 del req
.headers
['Ytdl-socks-proxy']
1347 return self
.do_open(functools
.partial(
1348 _create_http_connection
, self
, conn_class
, False),
1356 return zlib
.decompress(data
, -zlib
.MAX_WBITS
)
1358 return zlib
.decompress(data
)
1360 def http_request(self
, req
):
1361 # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not
1362 # always respected by websites, some tend to give out URLs with non percent-encoded
1363 # non-ASCII characters (see telemb.py, ard.py [#3412])
1364 # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991)
1365 # To work around aforementioned issue we will replace request's original URL with
1366 # percent-encoded one
1367 # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09)
1368 # the code of this workaround has been moved here from YoutubeDL.urlopen()
1369 url
= req
.get_full_url()
1370 url_escaped
= escape_url(url
)
1372 # Substitute URL if any change after escaping
1373 if url
!= url_escaped
:
1374 req
= update_Request(req
, url
=url_escaped
)
1376 for h
, v
in std_headers
.items():
1377 # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275
1378 # The dict keys are capitalized because of this bug by urllib
1379 if h
.capitalize() not in req
.headers
:
1380 req
.add_header(h
, v
)
1382 req
.headers
= handle_youtubedl_headers(req
.headers
)
1384 if sys
.version_info
< (2, 7) and '#' in req
.get_full_url():
1385 # Python 2.6 is brain-dead when it comes to fragments
1386 req
._Request
__original
= req
._Request
__original
.partition('#')[0]
1387 req
._Request
__r
_type
= req
._Request
__r
_type
.partition('#')[0]
1391 def http_response(self
, req
, resp
):
1394 if resp
.headers
.get('Content-encoding', '') == 'gzip':
1395 content
= resp
.read()
1396 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
), mode
='rb')
1398 uncompressed
= io
.BytesIO(gz
.read())
1399 except IOError as original_ioerror
:
1400 # There may be junk add the end of the file
1401 # See http://stackoverflow.com/q/4928560/35070 for details
1402 for i
in range(1, 1024):
1404 gz
= gzip
.GzipFile(fileobj
=io
.BytesIO(content
[:-i
]), mode
='rb')
1405 uncompressed
= io
.BytesIO(gz
.read())
1410 raise original_ioerror
1411 resp
= compat_urllib_request
.addinfourl(uncompressed
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1412 resp
.msg
= old_resp
.msg
1413 del resp
.headers
['Content-encoding']
1415 if resp
.headers
.get('Content-encoding', '') == 'deflate':
1416 gz
= io
.BytesIO(self
.deflate(resp
.read()))
1417 resp
= compat_urllib_request
.addinfourl(gz
, old_resp
.headers
, old_resp
.url
, old_resp
.code
)
1418 resp
.msg
= old_resp
.msg
1419 del resp
.headers
['Content-encoding']
1420 # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 (see
1421 # https://github.com/ytdl-org/youtube-dl/issues/6457).
1422 if 300 <= resp
.code
< 400:
1423 location
= resp
.headers
.get('Location')
1425 # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3
1426 if sys
.version_info
>= (3, 0):
1427 location
= location
.encode('iso-8859-1').decode('utf-8')
1429 location
= location
.decode('utf-8')
1430 location_escaped
= escape_url(location
)
1431 if location
!= location_escaped
:
1432 del resp
.headers
['Location']
1433 if sys
.version_info
< (3, 0):
1434 location_escaped
= location_escaped
.encode('utf-8')
1435 resp
.headers
['Location'] = location_escaped
1438 https_request
= http_request
1439 https_response
= http_response
1442 def make_socks_conn_class(base_class
, socks_proxy
):
1443 assert issubclass(base_class
, (
1444 compat_http_client
.HTTPConnection
, compat_http_client
.HTTPSConnection
))
1446 url_components
= compat_urlparse
.urlparse(socks_proxy
)
1447 if url_components
.scheme
.lower() == 'socks5':
1448 socks_type
= ProxyType
.SOCKS5
1449 elif url_components
.scheme
.lower() in ('socks', 'socks4'):
1450 socks_type
= ProxyType
.SOCKS4
1451 elif url_components
.scheme
.lower() == 'socks4a':
1452 socks_type
= ProxyType
.SOCKS4A
1454 def unquote_if_non_empty(s
):
1457 return compat_urllib_parse_unquote_plus(s
)
1461 url_components
.hostname
, url_components
.port
or 1080,
1463 unquote_if_non_empty(url_components
.username
),
1464 unquote_if_non_empty(url_components
.password
),
1467 class SocksConnection(base_class
):
1469 self
.sock
= sockssocket()
1470 self
.sock
.setproxy(*proxy_args
)
1471 if type(self
.timeout
) in (int, float):
1472 self
.sock
.settimeout(self
.timeout
)
1473 self
.sock
.connect((self
.host
, self
.port
))
1475 if isinstance(self
, compat_http_client
.HTTPSConnection
):
1476 if hasattr(self
, '_context'): # Python > 2.6
1477 self
.sock
= self
._context
.wrap_socket(
1478 self
.sock
, server_hostname
=self
.host
)
1480 self
.sock
= ssl
.wrap_socket(self
.sock
)
1482 return SocksConnection
1485 class YoutubeDLHTTPSHandler(compat_urllib_request
.HTTPSHandler
):
1486 def __init__(self
, params
, https_conn_class
=None, *args
, **kwargs
):
1487 compat_urllib_request
.HTTPSHandler
.__init
__(self
, *args
, **kwargs
)
1488 self
._https
_conn
_class
= https_conn_class
or compat_http_client
.HTTPSConnection
1489 self
._params
= params
1491 def https_open(self
, req
):
1493 conn_class
= self
._https
_conn
_class
1495 if hasattr(self
, '_context'): # python > 2.6
1496 kwargs
['context'] = self
._context
1497 if hasattr(self
, '_check_hostname'): # python 3.x
1498 kwargs
['check_hostname'] = self
._check
_hostname
1500 socks_proxy
= req
.headers
.get('Ytdl-socks-proxy')
1502 conn_class
= make_socks_conn_class(conn_class
, socks_proxy
)
1503 del req
.headers
['Ytdl-socks-proxy']
1505 return self
.do_open(functools
.partial(
1506 _create_http_connection
, self
, conn_class
, True),
1510 class YoutubeDLCookieJar(compat_cookiejar
.MozillaCookieJar
):
1512 See [1] for cookie file format.
1514 1. https://curl.haxx.se/docs/http-cookies.html
1516 _HTTPONLY_PREFIX
= '#HttpOnly_'
1518 _HEADER
= '''# Netscape HTTP Cookie File
1519 # This file is generated by yt-dlp. Do not edit.
1522 _CookieFileEntry
= collections
.namedtuple(
1524 ('domain_name', 'include_subdomains', 'path', 'https_only', 'expires_at', 'name', 'value'))
1526 def save(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1528 Save cookies to a file.
1530 Most of the code is taken from CPython 3.8 and slightly adapted
1531 to support cookie files with UTF-8 in both python 2 and 3.
1533 if filename
is None:
1534 if self
.filename
is not None:
1535 filename
= self
.filename
1537 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1539 # Store session cookies with `expires` set to 0 instead of an empty
1542 if cookie
.expires
is None:
1545 with io
.open(filename
, 'w', encoding
='utf-8') as f
:
1546 f
.write(self
._HEADER
)
1549 if not ignore_discard
and cookie
.discard
:
1551 if not ignore_expires
and cookie
.is_expired(now
):
1557 if cookie
.domain
.startswith('.'):
1558 initial_dot
= 'TRUE'
1560 initial_dot
= 'FALSE'
1561 if cookie
.expires
is not None:
1562 expires
= compat_str(cookie
.expires
)
1565 if cookie
.value
is None:
1566 # cookies.txt regards 'Set-Cookie: foo' as a cookie
1567 # with no name, whereas http.cookiejar regards it as a
1568 # cookie with no value.
1573 value
= cookie
.value
1575 '\t'.join([cookie
.domain
, initial_dot
, cookie
.path
,
1576 secure
, expires
, name
, value
]) + '\n')
1578 def load(self
, filename
=None, ignore_discard
=False, ignore_expires
=False):
1579 """Load cookies from a file."""
1580 if filename
is None:
1581 if self
.filename
is not None:
1582 filename
= self
.filename
1584 raise ValueError(compat_cookiejar
.MISSING_FILENAME_TEXT
)
1586 def prepare_line(line
):
1587 if line
.startswith(self
._HTTPONLY
_PREFIX
):
1588 line
= line
[len(self
._HTTPONLY
_PREFIX
):]
1589 # comments and empty lines are fine
1590 if line
.startswith('#') or not line
.strip():
1592 cookie_list
= line
.split('\t')
1593 if len(cookie_list
) != self
._ENTRY
_LEN
:
1594 raise compat_cookiejar
.LoadError('invalid length %d' % len(cookie_list
))
1595 cookie
= self
._CookieFileEntry
(*cookie_list
)
1596 if cookie
.expires_at
and not cookie
.expires_at
.isdigit():
1597 raise compat_cookiejar
.LoadError('invalid expires at %s' % cookie
.expires_at
)
1601 with io
.open(filename
, encoding
='utf-8') as f
:
1604 cf
.write(prepare_line(line
))
1605 except compat_cookiejar
.LoadError
as e
:
1607 'WARNING: skipping cookie file entry due to %s: %r\n'
1608 % (e
, line
), sys
.stderr
)
1611 self
._really
_load
(cf
, filename
, ignore_discard
, ignore_expires
)
1612 # Session cookies are denoted by either `expires` field set to
1613 # an empty string or 0. MozillaCookieJar only recognizes the former
1614 # (see [1]). So we need force the latter to be recognized as session
1615 # cookies on our own.
1616 # Session cookies may be important for cookies-based authentication,
1617 # e.g. usually, when user does not check 'Remember me' check box while
1618 # logging in on a site, some important cookies are stored as session
1619 # cookies so that not recognizing them will result in failed login.
1620 # 1. https://bugs.python.org/issue17164
1622 # Treat `expires=0` cookies as session cookies
1623 if cookie
.expires
== 0:
1624 cookie
.expires
= None
1625 cookie
.discard
= True
1628 class YoutubeDLCookieProcessor(compat_urllib_request
.HTTPCookieProcessor
):
1629 def __init__(self
, cookiejar
=None):
1630 compat_urllib_request
.HTTPCookieProcessor
.__init
__(self
, cookiejar
)
1632 def http_response(self
, request
, response
):
1633 # Python 2 will choke on next HTTP request in row if there are non-ASCII
1634 # characters in Set-Cookie HTTP header of last response (see
1635 # https://github.com/ytdl-org/youtube-dl/issues/6769).
1636 # In order to at least prevent crashing we will percent encode Set-Cookie
1637 # header before HTTPCookieProcessor starts processing it.
1638 # if sys.version_info < (3, 0) and response.headers:
1639 # for set_cookie_header in ('Set-Cookie', 'Set-Cookie2'):
1640 # set_cookie = response.headers.get(set_cookie_header)
1642 # set_cookie_escaped = compat_urllib_parse.quote(set_cookie, b"%/;:@&=+$,!~*'()?#[] ")
1643 # if set_cookie != set_cookie_escaped:
1644 # del response.headers[set_cookie_header]
1645 # response.headers[set_cookie_header] = set_cookie_escaped
1646 return compat_urllib_request
.HTTPCookieProcessor
.http_response(self
, request
, response
)
1648 https_request
= compat_urllib_request
.HTTPCookieProcessor
.http_request
1649 https_response
= http_response
1652 class YoutubeDLRedirectHandler(compat_urllib_request
.HTTPRedirectHandler
):
1653 """YoutubeDL redirect handler
1655 The code is based on HTTPRedirectHandler implementation from CPython [1].
1657 This redirect handler solves two issues:
1658 - ensures redirect URL is always unicode under python 2
1659 - introduces support for experimental HTTP response status code
1660 308 Permanent Redirect [2] used by some sites [3]
1662 1. https://github.com/python/cpython/blob/master/Lib/urllib/request.py
1663 2. https://developer.mozilla.org/en-US/docs/Web/HTTP/Status/308
1664 3. https://github.com/ytdl-org/youtube-dl/issues/28768
1667 http_error_301
= http_error_303
= http_error_307
= http_error_308
= compat_urllib_request
.HTTPRedirectHandler
.http_error_302
1669 def redirect_request(self
, req
, fp
, code
, msg
, headers
, newurl
):
1670 """Return a Request or None in response to a redirect.
1672 This is called by the http_error_30x methods when a
1673 redirection response is received. If a redirection should
1674 take place, return a new Request to allow http_error_30x to
1675 perform the redirect. Otherwise, raise HTTPError if no-one
1676 else should try to handle this url. Return None if you can't
1677 but another Handler might.
1679 m
= req
.get_method()
1680 if (not (code
in (301, 302, 303, 307, 308) and m
in ("GET", "HEAD")
1681 or code
in (301, 302, 303) and m
== "POST")):
1682 raise compat_HTTPError(req
.full_url
, code
, msg
, headers
, fp
)
1683 # Strictly (according to RFC 2616), 301 or 302 in response to
1684 # a POST MUST NOT cause a redirection without confirmation
1685 # from the user (of urllib.request, in this case). In practice,
1686 # essentially all clients do redirect in this case, so we do
1689 # On python 2 urlh.geturl() may sometimes return redirect URL
1690 # as byte string instead of unicode. This workaround allows
1691 # to force it always return unicode.
1692 if sys
.version_info
[0] < 3:
1693 newurl
= compat_str(newurl
)
1695 # Be conciliant with URIs containing a space. This is mainly
1696 # redundant with the more complete encoding done in http_error_302(),
1697 # but it is kept for compatibility with other callers.
1698 newurl
= newurl
.replace(' ', '%20')
1700 CONTENT_HEADERS
= ("content-length", "content-type")
1701 # NB: don't use dict comprehension for python 2.6 compatibility
1702 newheaders
= dict((k
, v
) for k
, v
in req
.headers
.items()
1703 if k
.lower() not in CONTENT_HEADERS
)
1704 return compat_urllib_request
.Request(
1705 newurl
, headers
=newheaders
, origin_req_host
=req
.origin_req_host
,
1709 def extract_timezone(date_str
):
1712 ^.{8,}? # >=8 char non-TZ prefix, if present
1713 (?P<tz>Z| # just the UTC Z, or
1714 (?:(?<=.\b\d{4}|\b\d{2}:\d\d)| # preceded by 4 digits or hh:mm or
1715 (?<!.\b[a-zA-Z]{3}|[a-zA-Z]{4}|..\b\d\d)) # not preceded by 3 alpha word or >= 4 alpha or 2 digits
1716 [ ]? # optional space
1717 (?P<sign>\+|-) # +/-
1718 (?P<hours>[0-9]{2}):?(?P<minutes>[0-9]{2}) # hh[:]mm
1722 timezone
= datetime
.timedelta()
1724 date_str
= date_str
[:-len(m
.group('tz'))]
1725 if not m
.group('sign'):
1726 timezone
= datetime
.timedelta()
1728 sign
= 1 if m
.group('sign') == '+' else -1
1729 timezone
= datetime
.timedelta(
1730 hours
=sign
* int(m
.group('hours')),
1731 minutes
=sign
* int(m
.group('minutes')))
1732 return timezone
, date_str
1735 def parse_iso8601(date_str
, delimiter
='T', timezone
=None):
1736 """ Return a UNIX timestamp from the given date """
1738 if date_str
is None:
1741 date_str
= re
.sub(r
'\.[0-9]+', '', date_str
)
1743 if timezone
is None:
1744 timezone
, date_str
= extract_timezone(date_str
)
1747 date_format
= '%Y-%m-%d{0}%H:%M:%S'.format(delimiter
)
1748 dt
= datetime
.datetime
.strptime(date_str
, date_format
) - timezone
1749 return calendar
.timegm(dt
.timetuple())
1754 def date_formats(day_first
=True):
1755 return DATE_FORMATS_DAY_FIRST
if day_first
else DATE_FORMATS_MONTH_FIRST
1758 def unified_strdate(date_str
, day_first
=True):
1759 """Return a string with the date in the format YYYYMMDD"""
1761 if date_str
is None:
1765 date_str
= date_str
.replace(',', ' ')
1766 # Remove AM/PM + timezone
1767 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1768 _
, date_str
= extract_timezone(date_str
)
1770 for expression
in date_formats(day_first
):
1772 upload_date
= datetime
.datetime
.strptime(date_str
, expression
).strftime('%Y%m%d')
1775 if upload_date
is None:
1776 timetuple
= email
.utils
.parsedate_tz(date_str
)
1779 upload_date
= datetime
.datetime(*timetuple
[:6]).strftime('%Y%m%d')
1782 if upload_date
is not None:
1783 return compat_str(upload_date
)
1786 def unified_timestamp(date_str
, day_first
=True):
1787 if date_str
is None:
1790 date_str
= re
.sub(r
'[,|]', '', date_str
)
1792 pm_delta
= 12 if re
.search(r
'(?i)PM', date_str
) else 0
1793 timezone
, date_str
= extract_timezone(date_str
)
1795 # Remove AM/PM + timezone
1796 date_str
= re
.sub(r
'(?i)\s*(?:AM|PM)(?:\s+[A-Z]+)?', '', date_str
)
1798 # Remove unrecognized timezones from ISO 8601 alike timestamps
1799 m
= re
.search(r
'\d{1,2}:\d{1,2}(?:\.\d+)?(?P<tz>\s*[A-Z]+)$', date_str
)
1801 date_str
= date_str
[:-len(m
.group('tz'))]
1803 # Python only supports microseconds, so remove nanoseconds
1804 m
= re
.search(r
'^([0-9]{4,}-[0-9]{1,2}-[0-9]{1,2}T[0-9]{1,2}:[0-9]{1,2}:[0-9]{1,2}\.[0-9]{6})[0-9]+$', date_str
)
1806 date_str
= m
.group(1)
1808 for expression
in date_formats(day_first
):
1810 dt
= datetime
.datetime
.strptime(date_str
, expression
) - timezone
+ datetime
.timedelta(hours
=pm_delta
)
1811 return calendar
.timegm(dt
.timetuple())
1814 timetuple
= email
.utils
.parsedate_tz(date_str
)
1816 return calendar
.timegm(timetuple
) + pm_delta
* 3600
1819 def determine_ext(url
, default_ext
='unknown_video'):
1820 if url
is None or '.' not in url
:
1822 guess
= url
.partition('?')[0].rpartition('.')[2]
1823 if re
.match(r
'^[A-Za-z0-9]+$', guess
):
1825 # Try extract ext from URLs like http://example.com/foo/bar.mp4/?download
1826 elif guess
.rstrip('/') in KNOWN_EXTENSIONS
:
1827 return guess
.rstrip('/')
1832 def subtitles_filename(filename
, sub_lang
, sub_format
, expected_real_ext
=None):
1833 return replace_extension(filename
, sub_lang
+ '.' + sub_format
, expected_real_ext
)
1836 def datetime_from_str(date_str
, precision
='auto', format
='%Y%m%d'):
1838 Return a datetime object from a string in the format YYYYMMDD or
1839 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1841 format: string date format used to return datetime object from
1842 precision: round the time portion of a datetime object.
1843 auto|microsecond|second|minute|hour|day.
1844 auto: round to the unit provided in date_str (if applicable).
1846 auto_precision
= False
1847 if precision
== 'auto':
1848 auto_precision
= True
1849 precision
= 'microsecond'
1850 today
= datetime_round(datetime
.datetime
.utcnow(), precision
)
1851 if date_str
in ('now', 'today'):
1853 if date_str
== 'yesterday':
1854 return today
- datetime
.timedelta(days
=1)
1856 r
'(?P<start>.+)(?P<sign>[+-])(?P<time>\d+)(?P<unit>microsecond|second|minute|hour|day|week|month|year)(s)?',
1858 if match
is not None:
1859 start_time
= datetime_from_str(match
.group('start'), precision
, format
)
1860 time
= int(match
.group('time')) * (-1 if match
.group('sign') == '-' else 1)
1861 unit
= match
.group('unit')
1862 if unit
== 'month' or unit
== 'year':
1863 new_date
= datetime_add_months(start_time
, time
* 12 if unit
== 'year' else time
)
1869 delta
= datetime
.timedelta(**{unit + 's': time}
)
1870 new_date
= start_time
+ delta
1872 return datetime_round(new_date
, unit
)
1875 return datetime_round(datetime
.datetime
.strptime(date_str
, format
), precision
)
1878 def date_from_str(date_str
, format
='%Y%m%d'):
1880 Return a datetime object from a string in the format YYYYMMDD or
1881 (now|today|date)[+-][0-9](microsecond|second|minute|hour|day|week|month|year)(s)?
1883 format: string date format used to return datetime object from
1885 return datetime_from_str(date_str
, precision
='microsecond', format
=format
).date()
1888 def datetime_add_months(dt
, months
):
1889 """Increment/Decrement a datetime object by months."""
1890 month
= dt
.month
+ months
- 1
1891 year
= dt
.year
+ month
// 12
1892 month
= month
% 12 + 1
1893 day
= min(dt
.day
, calendar
.monthrange(year
, month
)[1])
1894 return dt
.replace(year
, month
, day
)
1897 def datetime_round(dt
, precision
='day'):
1899 Round a datetime object's time to a specific precision
1901 if precision
== 'microsecond':
1910 roundto
= lambda x
, n
: ((x
+ n
/ 2) // n
) * n
1911 timestamp
= calendar
.timegm(dt
.timetuple())
1912 return datetime
.datetime
.utcfromtimestamp(roundto(timestamp
, unit_seconds
[precision
]))
1915 def hyphenate_date(date_str
):
1917 Convert a date in 'YYYYMMDD' format to 'YYYY-MM-DD' format"""
1918 match
= re
.match(r
'^(\d\d\d\d)(\d\d)(\d\d)$', date_str
)
1919 if match
is not None:
1920 return '-'.join(match
.groups())
1925 class DateRange(object):
1926 """Represents a time interval between two dates"""
1928 def __init__(self
, start
=None, end
=None):
1929 """start and end must be strings in the format accepted by date"""
1930 if start
is not None:
1931 self
.start
= date_from_str(start
)
1933 self
.start
= datetime
.datetime
.min.date()
1935 self
.end
= date_from_str(end
)
1937 self
.end
= datetime
.datetime
.max.date()
1938 if self
.start
> self
.end
:
1939 raise ValueError('Date range: "%s" , the start date must be before the end date' % self
)
1943 """Returns a range that only contains the given day"""
1944 return cls(day
, day
)
1946 def __contains__(self
, date
):
1947 """Check if the date is in the range"""
1948 if not isinstance(date
, datetime
.date
):
1949 date
= date_from_str(date
)
1950 return self
.start
<= date
<= self
.end
1953 return '%s - %s' % (self
.start
.isoformat(), self
.end
.isoformat())
1956 def platform_name():
1957 """ Returns the platform name as a compat_str """
1958 res
= platform
.platform()
1959 if isinstance(res
, bytes):
1960 res
= res
.decode(preferredencoding())
1962 assert isinstance(res
, compat_str
)
1966 def get_windows_version():
1967 ''' Get Windows version. None if it's not running on Windows '''
1968 if compat_os_name
== 'nt':
1969 return version_tuple(platform
.win32_ver()[1])
1974 def _windows_write_string(s
, out
):
1975 """ Returns True if the string was written using special methods,
1976 False if it has yet to be written out."""
1977 # Adapted from http://stackoverflow.com/a/3259271/35070
1979 import ctypes
.wintypes
1987 fileno
= out
.fileno()
1988 except AttributeError:
1989 # If the output stream doesn't have a fileno, it's virtual
1991 except io
.UnsupportedOperation
:
1992 # Some strange Windows pseudo files?
1994 if fileno
not in WIN_OUTPUT_IDS
:
1997 GetStdHandle
= compat_ctypes_WINFUNCTYPE(
1998 ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.DWORD
)(
1999 ('GetStdHandle', ctypes
.windll
.kernel32
))
2000 h
= GetStdHandle(WIN_OUTPUT_IDS
[fileno
])
2002 WriteConsoleW
= compat_ctypes_WINFUNCTYPE(
2003 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
, ctypes
.wintypes
.LPWSTR
,
2004 ctypes
.wintypes
.DWORD
, ctypes
.POINTER(ctypes
.wintypes
.DWORD
),
2005 ctypes
.wintypes
.LPVOID
)(('WriteConsoleW', ctypes
.windll
.kernel32
))
2006 written
= ctypes
.wintypes
.DWORD(0)
2008 GetFileType
= compat_ctypes_WINFUNCTYPE(ctypes
.wintypes
.DWORD
, ctypes
.wintypes
.DWORD
)(('GetFileType', ctypes
.windll
.kernel32
))
2009 FILE_TYPE_CHAR
= 0x0002
2010 FILE_TYPE_REMOTE
= 0x8000
2011 GetConsoleMode
= compat_ctypes_WINFUNCTYPE(
2012 ctypes
.wintypes
.BOOL
, ctypes
.wintypes
.HANDLE
,
2013 ctypes
.POINTER(ctypes
.wintypes
.DWORD
))(
2014 ('GetConsoleMode', ctypes
.windll
.kernel32
))
2015 INVALID_HANDLE_VALUE
= ctypes
.wintypes
.DWORD(-1).value
2017 def not_a_console(handle
):
2018 if handle
== INVALID_HANDLE_VALUE
or handle
is None:
2020 return ((GetFileType(handle
) & ~FILE_TYPE_REMOTE
) != FILE_TYPE_CHAR
2021 or GetConsoleMode(handle
, ctypes
.byref(ctypes
.wintypes
.DWORD())) == 0)
2023 if not_a_console(h
):
2026 def next_nonbmp_pos(s
):
2028 return next(i
for i
, c
in enumerate(s
) if ord(c
) > 0xffff)
2029 except StopIteration:
2033 count
= min(next_nonbmp_pos(s
), 1024)
2035 ret
= WriteConsoleW(
2036 h
, s
, count
if count
else 2, ctypes
.byref(written
), None)
2038 raise OSError('Failed to write string')
2039 if not count
: # We just wrote a non-BMP character
2040 assert written
.value
== 2
2043 assert written
.value
> 0
2044 s
= s
[written
.value
:]
2048 def write_string(s
, out
=None, encoding
=None):
2051 assert type(s
) == compat_str
2053 if sys
.platform
== 'win32' and encoding
is None and hasattr(out
, 'fileno'):
2054 if _windows_write_string(s
, out
):
2057 if ('b' in getattr(out
, 'mode', '')
2058 or sys
.version_info
[0] < 3): # Python 2 lies about mode of sys.stderr
2059 byt
= s
.encode(encoding
or preferredencoding(), 'ignore')
2061 elif hasattr(out
, 'buffer'):
2062 enc
= encoding
or getattr(out
, 'encoding', None) or preferredencoding()
2063 byt
= s
.encode(enc
, 'ignore')
2064 out
.buffer.write(byt
)
2070 def bytes_to_intlist(bs
):
2073 if isinstance(bs
[0], int): # Python 3
2076 return [ord(c
) for c
in bs
]
2079 def intlist_to_bytes(xs
):
2082 return compat_struct_pack('%dB' % len(xs
), *xs
)
2085 # Cross-platform file locking
2086 if sys
.platform
== 'win32':
2087 import ctypes
.wintypes
2090 class OVERLAPPED(ctypes
.Structure
):
2092 ('Internal', ctypes
.wintypes
.LPVOID
),
2093 ('InternalHigh', ctypes
.wintypes
.LPVOID
),
2094 ('Offset', ctypes
.wintypes
.DWORD
),
2095 ('OffsetHigh', ctypes
.wintypes
.DWORD
),
2096 ('hEvent', ctypes
.wintypes
.HANDLE
),
2099 kernel32
= ctypes
.windll
.kernel32
2100 LockFileEx
= kernel32
.LockFileEx
2101 LockFileEx
.argtypes
= [
2102 ctypes
.wintypes
.HANDLE
, # hFile
2103 ctypes
.wintypes
.DWORD
, # dwFlags
2104 ctypes
.wintypes
.DWORD
, # dwReserved
2105 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2106 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2107 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2109 LockFileEx
.restype
= ctypes
.wintypes
.BOOL
2110 UnlockFileEx
= kernel32
.UnlockFileEx
2111 UnlockFileEx
.argtypes
= [
2112 ctypes
.wintypes
.HANDLE
, # hFile
2113 ctypes
.wintypes
.DWORD
, # dwReserved
2114 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockLow
2115 ctypes
.wintypes
.DWORD
, # nNumberOfBytesToLockHigh
2116 ctypes
.POINTER(OVERLAPPED
) # Overlapped
2118 UnlockFileEx
.restype
= ctypes
.wintypes
.BOOL
2119 whole_low
= 0xffffffff
2120 whole_high
= 0x7fffffff
2122 def _lock_file(f
, exclusive
):
2123 overlapped
= OVERLAPPED()
2124 overlapped
.Offset
= 0
2125 overlapped
.OffsetHigh
= 0
2126 overlapped
.hEvent
= 0
2127 f
._lock
_file
_overlapped
_p
= ctypes
.pointer(overlapped
)
2128 handle
= msvcrt
.get_osfhandle(f
.fileno())
2129 if not LockFileEx(handle
, 0x2 if exclusive
else 0x0, 0,
2130 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2131 raise OSError('Locking file failed: %r' % ctypes
.FormatError())
2133 def _unlock_file(f
):
2134 assert f
._lock
_file
_overlapped
_p
2135 handle
= msvcrt
.get_osfhandle(f
.fileno())
2136 if not UnlockFileEx(handle
, 0,
2137 whole_low
, whole_high
, f
._lock
_file
_overlapped
_p
):
2138 raise OSError('Unlocking file failed: %r' % ctypes
.FormatError())
2141 # Some platforms, such as Jython, is missing fcntl
2145 def _lock_file(f
, exclusive
):
2146 fcntl
.flock(f
, fcntl
.LOCK_EX
if exclusive
else fcntl
.LOCK_SH
)
2148 def _unlock_file(f
):
2149 fcntl
.flock(f
, fcntl
.LOCK_UN
)
2151 UNSUPPORTED_MSG
= 'file locking is not supported on this platform'
2153 def _lock_file(f
, exclusive
):
2154 raise IOError(UNSUPPORTED_MSG
)
2156 def _unlock_file(f
):
2157 raise IOError(UNSUPPORTED_MSG
)
2160 class locked_file(object):
2161 def __init__(self
, filename
, mode
, encoding
=None):
2162 assert mode
in ['r', 'a', 'w']
2163 self
.f
= io
.open(filename
, mode
, encoding
=encoding
)
2166 def __enter__(self
):
2167 exclusive
= self
.mode
!= 'r'
2169 _lock_file(self
.f
, exclusive
)
2175 def __exit__(self
, etype
, value
, traceback
):
2177 _unlock_file(self
.f
)
2184 def write(self
, *args
):
2185 return self
.f
.write(*args
)
2187 def read(self
, *args
):
2188 return self
.f
.read(*args
)
2191 def get_filesystem_encoding():
2192 encoding
= sys
.getfilesystemencoding()
2193 return encoding
if encoding
is not None else 'utf-8'
2196 def shell_quote(args
):
2198 encoding
= get_filesystem_encoding()
2200 if isinstance(a
, bytes):
2201 # We may get a filename encoded with 'encodeFilename'
2202 a
= a
.decode(encoding
)
2203 quoted_args
.append(compat_shlex_quote(a
))
2204 return ' '.join(quoted_args
)
2207 def smuggle_url(url
, data
):
2208 """ Pass additional data in a URL for internal use. """
2210 url
, idata
= unsmuggle_url(url
, {})
2212 sdata
= compat_urllib_parse_urlencode(
2213 {'__youtubedl_smuggle': json.dumps(data)}
)
2214 return url
+ '#' + sdata
2217 def unsmuggle_url(smug_url
, default
=None):
2218 if '#__youtubedl_smuggle' not in smug_url
:
2219 return smug_url
, default
2220 url
, _
, sdata
= smug_url
.rpartition('#')
2221 jsond
= compat_parse_qs(sdata
)['__youtubedl_smuggle'][0]
2222 data
= json
.loads(jsond
)
2226 def format_decimal_suffix(num
, fmt
='%d%s', *, factor
=1000):
2227 """ Formats numbers with decimal sufixes like K, M, etc """
2228 num
, factor
= float_or_none(num
), float(factor
)
2231 exponent
= 0 if num
== 0 else int(math
.log(num
, factor
))
2232 suffix
= ['', *'kMGTPEZY'][exponent
]
2234 suffix
= {'k': 'Ki', '': ''}
.get(suffix
, f
'{suffix}i')
2235 converted
= num
/ (factor
** exponent
)
2236 return fmt
% (converted
, suffix
)
2239 def format_bytes(bytes):
2240 return format_decimal_suffix(bytes, '%.2f%sB', factor
=1024) or 'N/A'
2243 def lookup_unit_table(unit_table
, s
):
2244 units_re
= '|'.join(re
.escape(u
) for u
in unit_table
)
2246 r
'(?P<num>[0-9]+(?:[,.][0-9]*)?)\s*(?P<unit>%s)\b' % units_re
, s
)
2249 num_str
= m
.group('num').replace(',', '.')
2250 mult
= unit_table
[m
.group('unit')]
2251 return int(float(num_str
) * mult
)
2254 def parse_filesize(s
):
2258 # The lower-case forms are of course incorrect and unofficial,
2259 # but we support those too
2276 'megabytes': 1000 ** 2,
2277 'mebibytes': 1024 ** 2,
2283 'gigabytes': 1000 ** 3,
2284 'gibibytes': 1024 ** 3,
2290 'terabytes': 1000 ** 4,
2291 'tebibytes': 1024 ** 4,
2297 'petabytes': 1000 ** 5,
2298 'pebibytes': 1024 ** 5,
2304 'exabytes': 1000 ** 6,
2305 'exbibytes': 1024 ** 6,
2311 'zettabytes': 1000 ** 7,
2312 'zebibytes': 1024 ** 7,
2318 'yottabytes': 1000 ** 8,
2319 'yobibytes': 1024 ** 8,
2322 return lookup_unit_table(_UNIT_TABLE
, s
)
2329 s
= re
.sub(r
'^[^\d]+\s', '', s
).strip()
2331 if re
.match(r
'^[\d,.]+$', s
):
2332 return str_to_int(s
)
2345 ret
= lookup_unit_table(_UNIT_TABLE
, s
)
2349 mobj
= re
.match(r
'([\d,.]+)(?:$|\s)', s
)
2351 return str_to_int(mobj
.group(1))
2354 def parse_resolution(s
):
2358 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(?P<w>\d+)\s*[xX×,]\s*(?P<h>\d+)(?![a-zA-Z0-9])', s
)
2361 'width': int(mobj
.group('w')),
2362 'height': int(mobj
.group('h')),
2365 mobj
= re
.search(r
'(?<![a-zA-Z0-9])(\d+)[pPiI](?![a-zA-Z0-9])', s
)
2367 return {'height': int(mobj.group(1))}
2369 mobj
= re
.search(r
'\b([48])[kK]\b', s
)
2371 return {'height': int(mobj.group(1)) * 540}
2376 def parse_bitrate(s
):
2377 if not isinstance(s
, compat_str
):
2379 mobj
= re
.search(r
'\b(\d+)\s*kbps', s
)
2381 return int(mobj
.group(1))
2384 def month_by_name(name
, lang
='en'):
2385 """ Return the number of a month by (locale-independently) English name """
2387 month_names
= MONTH_NAMES
.get(lang
, MONTH_NAMES
['en'])
2390 return month_names
.index(name
) + 1
2395 def month_by_abbreviation(abbrev
):
2396 """ Return the number of a month by (locale-independently) English
2400 return [s
[:3] for s
in ENGLISH_MONTH_NAMES
].index(abbrev
) + 1
2405 def fix_xml_ampersands(xml_str
):
2406 """Replace all the '&' by '&' in XML"""
2408 r
'&(?!amp;|lt;|gt;|apos;|quot;|#x[0-9a-fA-F]{,4};|#[0-9]{,4};)',
2413 def setproctitle(title
):
2414 assert isinstance(title
, compat_str
)
2416 # ctypes in Jython is not complete
2417 # http://bugs.jython.org/issue2148
2418 if sys
.platform
.startswith('java'):
2422 libc
= ctypes
.cdll
.LoadLibrary('libc.so.6')
2426 # LoadLibrary in Windows Python 2.7.13 only expects
2427 # a bytestring, but since unicode_literals turns
2428 # every string into a unicode string, it fails.
2430 title_bytes
= title
.encode('utf-8')
2431 buf
= ctypes
.create_string_buffer(len(title_bytes
))
2432 buf
.value
= title_bytes
2434 libc
.prctl(15, buf
, 0, 0, 0)
2435 except AttributeError:
2436 return # Strange libc, just skip this
2439 def remove_start(s
, start
):
2440 return s
[len(start
):] if s
is not None and s
.startswith(start
) else s
2443 def remove_end(s
, end
):
2444 return s
[:-len(end
)] if s
is not None and s
.endswith(end
) else s
2447 def remove_quotes(s
):
2448 if s
is None or len(s
) < 2:
2450 for quote
in ('"', "'", ):
2451 if s
[0] == quote
and s
[-1] == quote
:
2456 def get_domain(url
):
2457 domain
= re
.match(r
'(?:https?:\/\/)?(?:www\.)?(?P<domain>[^\n\/]+\.[^\n\/]+)(?:\/(.*))?', url
)
2458 return domain
.group('domain') if domain
else None
2461 def url_basename(url
):
2462 path
= compat_urlparse
.urlparse(url
).path
2463 return path
.strip('/').split('/')[-1]
2467 return re
.match(r
'https?://[^?#&]+/', url
).group()
2470 def urljoin(base
, path
):
2471 if isinstance(path
, bytes):
2472 path
= path
.decode('utf-8')
2473 if not isinstance(path
, compat_str
) or not path
:
2475 if re
.match(r
'^(?:[a-zA-Z][a-zA-Z0-9+-.]*:)?//', path
):
2477 if isinstance(base
, bytes):
2478 base
= base
.decode('utf-8')
2479 if not isinstance(base
, compat_str
) or not re
.match(
2480 r
'^(?:https?:)?//', base
):
2482 return compat_urlparse
.urljoin(base
, path
)
2485 class HEADRequest(compat_urllib_request
.Request
):
2486 def get_method(self
):
2490 class PUTRequest(compat_urllib_request
.Request
):
2491 def get_method(self
):
2495 def int_or_none(v
, scale
=1, default
=None, get_attr
=None, invscale
=1):
2496 if get_attr
and v
is not None:
2497 v
= getattr(v
, get_attr
, None)
2499 return int(v
) * invscale
// scale
2500 except (ValueError, TypeError, OverflowError):
2504 def str_or_none(v
, default
=None):
2505 return default
if v
is None else compat_str(v
)
2508 def str_to_int(int_str
):
2509 """ A more relaxed version of int_or_none """
2510 if isinstance(int_str
, compat_integer_types
):
2512 elif isinstance(int_str
, compat_str
):
2513 int_str
= re
.sub(r
'[,\.\+]', '', int_str
)
2514 return int_or_none(int_str
)
2517 def float_or_none(v
, scale
=1, invscale
=1, default
=None):
2521 return float(v
) * invscale
/ scale
2522 except (ValueError, TypeError):
2526 def bool_or_none(v
, default
=None):
2527 return v
if isinstance(v
, bool) else default
2530 def strip_or_none(v
, default
=None):
2531 return v
.strip() if isinstance(v
, compat_str
) else default
2534 def url_or_none(url
):
2535 if not url
or not isinstance(url
, compat_str
):
2538 return url
if re
.match(r
'^(?:(?:https?|rt(?:m(?:pt?[es]?|fp)|sp[su]?)|mms|ftps?):)?//', url
) else None
2541 def strftime_or_none(timestamp
, date_format
, default
=None):
2542 datetime_object
= None
2544 if isinstance(timestamp
, compat_numeric_types
): # unix timestamp
2545 datetime_object
= datetime
.datetime
.utcfromtimestamp(timestamp
)
2546 elif isinstance(timestamp
, compat_str
): # assume YYYYMMDD
2547 datetime_object
= datetime
.datetime
.strptime(timestamp
, '%Y%m%d')
2548 return datetime_object
.strftime(date_format
)
2549 except (ValueError, TypeError, AttributeError):
2553 def parse_duration(s
):
2554 if not isinstance(s
, compat_basestring
):
2560 days
, hours
, mins
, secs
, ms
= [None] * 5
2561 m
= re
.match(r
'''(?x)
2563 (?:(?:(?P<days>[0-9]+):)?(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?
2564 (?P<secs>(?(before_secs)[0-9]{1,2}|[0-9]+))
2565 (?P<ms>[.:][0-9]+)?Z?$
2568 days
, hours
, mins
, secs
, ms
= m
.group('days', 'hours', 'mins', 'secs', 'ms')
2573 [0-9]+\s*y(?:ears?)?\s*
2576 [0-9]+\s*m(?:onths?)?\s*
2579 [0-9]+\s*w(?:eeks?)?\s*
2582 (?P<days>[0-9]+)\s*d(?:ays?)?\s*
2586 (?P<hours>[0-9]+)\s*h(?:ours?)?\s*
2589 (?P<mins>[0-9]+)\s*m(?:in(?:ute)?s?)?\s*
2592 (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*s(?:ec(?:ond)?s?)?\s*
2595 days
, hours
, mins
, secs
, ms
= m
.groups()
2597 m
= re
.match(r
'(?i)(?:(?P<hours>[0-9.]+)\s*(?:hours?)|(?P<mins>[0-9.]+)\s*(?:mins?\.?|minutes?)\s*)Z?$', s
)
2599 hours
, mins
= m
.groups()
2605 duration
+= float(secs
)
2607 duration
+= float(mins
) * 60
2609 duration
+= float(hours
) * 60 * 60
2611 duration
+= float(days
) * 24 * 60 * 60
2613 duration
+= float(ms
.replace(':', '.'))
2617 def prepend_extension(filename
, ext
, expected_real_ext
=None):
2618 name
, real_ext
= os
.path
.splitext(filename
)
2620 '{0}.{1}{2}'.format(name
, ext
, real_ext
)
2621 if not expected_real_ext
or real_ext
[1:] == expected_real_ext
2622 else '{0}.{1}'.format(filename
, ext
))
2625 def replace_extension(filename
, ext
, expected_real_ext
=None):
2626 name
, real_ext
= os
.path
.splitext(filename
)
2627 return '{0}.{1}'.format(
2628 name
if not expected_real_ext
or real_ext
[1:] == expected_real_ext
else filename
,
2632 def check_executable(exe
, args
=[]):
2633 """ Checks if the given binary is installed somewhere in PATH, and returns its name.
2634 args can be a list of arguments for a short output (like -version) """
2636 Popen([exe
] + args
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
).communicate_or_kill()
2642 def _get_exe_version_output(exe
, args
):
2644 # STDIN should be redirected too. On UNIX-like systems, ffmpeg triggers
2645 # SIGTTOU if yt-dlp is run in the background.
2646 # See https://github.com/ytdl-org/youtube-dl/issues/955#issuecomment-209789656
2648 [encodeArgument(exe
)] + args
, stdin
=subprocess
.PIPE
,
2649 stdout
=subprocess
.PIPE
, stderr
=subprocess
.STDOUT
).communicate_or_kill()
2652 if isinstance(out
, bytes): # Python 2.x
2653 out
= out
.decode('ascii', 'ignore')
2657 def detect_exe_version(output
, version_re
=None, unrecognized
='present'):
2658 assert isinstance(output
, compat_str
)
2659 if version_re
is None:
2660 version_re
= r
'version\s+([-0-9._a-zA-Z]+)'
2661 m
= re
.search(version_re
, output
)
2668 def get_exe_version(exe
, args
=['--version'],
2669 version_re
=None, unrecognized
='present'):
2670 """ Returns the version of the specified executable,
2671 or False if the executable is not present """
2672 out
= _get_exe_version_output(exe
, args
)
2673 return detect_exe_version(out
, version_re
, unrecognized
) if out
else False
2676 class LazyList(collections
.abc
.Sequence
):
2677 ''' Lazy immutable list from an iterable
2678 Note that slices of a LazyList are lists and not LazyList'''
2680 class IndexError(IndexError):
2683 def __init__(self
, iterable
, *, reverse
=False, _cache
=None):
2684 self
.__iterable
= iter(iterable
)
2685 self
.__cache
= [] if _cache
is None else _cache
2686 self
.__reversed
= reverse
2690 # We need to consume the entire iterable to iterate in reverse
2691 yield from self
.exhaust()
2693 yield from self
.__cache
2694 for item
in self
.__iterable
:
2695 self
.__cache
.append(item
)
2698 def __exhaust(self
):
2699 self
.__cache
.extend(self
.__iterable
)
2700 # Discard the emptied iterable to make it pickle-able
2701 self
.__iterable
= []
2705 ''' Evaluate the entire iterable '''
2706 return self
.__exhaust
()[::-1 if self
.__reversed
else 1]
2709 def __reverse_index(x
):
2710 return None if x
is None else -(x
+ 1)
2712 def __getitem__(self
, idx
):
2713 if isinstance(idx
, slice):
2715 idx
= slice(self
.__reverse
_index
(idx
.start
), self
.__reverse
_index
(idx
.stop
), -(idx
.step
or 1))
2716 start
, stop
, step
= idx
.start
, idx
.stop
, idx
.step
or 1
2717 elif isinstance(idx
, int):
2719 idx
= self
.__reverse
_index
(idx
)
2720 start
, stop
, step
= idx
, idx
, 0
2722 raise TypeError('indices must be integers or slices')
2723 if ((start
or 0) < 0 or (stop
or 0) < 0
2724 or (start
is None and step
< 0)
2725 or (stop
is None and step
> 0)):
2726 # We need to consume the entire iterable to be able to slice from the end
2727 # Obviously, never use this with infinite iterables
2730 return self
.__cache
[idx
]
2731 except IndexError as e
:
2732 raise self
.IndexError(e
) from e
2733 n
= max(start
or 0, stop
or 0) - len(self
.__cache
) + 1
2735 self
.__cache
.extend(itertools
.islice(self
.__iterable
, n
))
2737 return self
.__cache
[idx
]
2738 except IndexError as e
:
2739 raise self
.IndexError(e
) from e
2743 self
[-1] if self
.__reversed
else self
[0]
2744 except self
.IndexError:
2750 return len(self
.__cache
)
2752 def __reversed__(self
):
2753 return type(self
)(self
.__iterable
, reverse
=not self
.__reversed
, _cache
=self
.__cache
)
2756 return type(self
)(self
.__iterable
, reverse
=self
.__reversed
, _cache
=self
.__cache
)
2759 # repr and str should mimic a list. So we exhaust the iterable
2760 return repr(self
.exhaust())
2763 return repr(self
.exhaust())
2768 class IndexError(IndexError):
2772 # This is only useful for tests
2773 return len(self
.getslice())
2775 def __init__(self
, pagefunc
, pagesize
, use_cache
=True):
2776 self
._pagefunc
= pagefunc
2777 self
._pagesize
= pagesize
2778 self
._use
_cache
= use_cache
2781 def getpage(self
, pagenum
):
2782 page_results
= self
._cache
.get(pagenum
)
2783 if page_results
is None:
2784 page_results
= list(self
._pagefunc
(pagenum
))
2786 self
._cache
[pagenum
] = page_results
2789 def getslice(self
, start
=0, end
=None):
2790 return list(self
._getslice
(start
, end
))
2792 def _getslice(self
, start
, end
):
2793 raise NotImplementedError('This method must be implemented by subclasses')
2795 def __getitem__(self
, idx
):
2796 # NOTE: cache must be enabled if this is used
2797 if not isinstance(idx
, int) or idx
< 0:
2798 raise TypeError('indices must be non-negative integers')
2799 entries
= self
.getslice(idx
, idx
+ 1)
2801 raise self
.IndexError()
2805 class OnDemandPagedList(PagedList
):
2806 def _getslice(self
, start
, end
):
2807 for pagenum
in itertools
.count(start
// self
._pagesize
):
2808 firstid
= pagenum
* self
._pagesize
2809 nextfirstid
= pagenum
* self
._pagesize
+ self
._pagesize
2810 if start
>= nextfirstid
:
2814 start
% self
._pagesize
2815 if firstid
<= start
< nextfirstid
2818 ((end
- 1) % self
._pagesize
) + 1
2819 if (end
is not None and firstid
<= end
<= nextfirstid
)
2822 page_results
= self
.getpage(pagenum
)
2823 if startv
!= 0 or endv
is not None:
2824 page_results
= page_results
[startv
:endv
]
2825 yield from page_results
2827 # A little optimization - if current page is not "full", ie. does
2828 # not contain page_size videos then we can assume that this page
2829 # is the last one - there are no more ids on further pages -
2830 # i.e. no need to query again.
2831 if len(page_results
) + startv
< self
._pagesize
:
2834 # If we got the whole page, but the next page is not interesting,
2835 # break out early as well
2836 if end
== nextfirstid
:
2840 class InAdvancePagedList(PagedList
):
2841 def __init__(self
, pagefunc
, pagecount
, pagesize
):
2842 self
._pagecount
= pagecount
2843 PagedList
.__init
__(self
, pagefunc
, pagesize
, True)
2845 def _getslice(self
, start
, end
):
2846 start_page
= start
// self
._pagesize
2847 end_page
= self
._pagecount
if end
is None else min(self
._pagecount
, end
// self
._pagesize
+ 1)
2848 skip_elems
= start
- start_page
* self
._pagesize
2849 only_more
= None if end
is None else end
- start
2850 for pagenum
in range(start_page
, end_page
):
2851 page_results
= self
.getpage(pagenum
)
2853 page_results
= page_results
[skip_elems
:]
2855 if only_more
is not None:
2856 if len(page_results
) < only_more
:
2857 only_more
-= len(page_results
)
2859 yield from page_results
[:only_more
]
2861 yield from page_results
2864 def uppercase_escape(s
):
2865 unicode_escape
= codecs
.getdecoder('unicode_escape')
2867 r
'\\U[0-9a-fA-F]{8}',
2868 lambda m
: unicode_escape(m
.group(0))[0],
2872 def lowercase_escape(s
):
2873 unicode_escape
= codecs
.getdecoder('unicode_escape')
2875 r
'\\u[0-9a-fA-F]{4}',
2876 lambda m
: unicode_escape(m
.group(0))[0],
2880 def escape_rfc3986(s
):
2881 """Escape non-ASCII characters as suggested by RFC 3986"""
2882 if sys
.version_info
< (3, 0) and isinstance(s
, compat_str
):
2883 s
= s
.encode('utf-8')
2884 return compat_urllib_parse
.quote(s
, b
"%/;:@&=+$,!~*'()?#[]")
2887 def escape_url(url
):
2888 """Escape URL as suggested by RFC 3986"""
2889 url_parsed
= compat_urllib_parse_urlparse(url
)
2890 return url_parsed
._replace
(
2891 netloc
=url_parsed
.netloc
.encode('idna').decode('ascii'),
2892 path
=escape_rfc3986(url_parsed
.path
),
2893 params
=escape_rfc3986(url_parsed
.params
),
2894 query
=escape_rfc3986(url_parsed
.query
),
2895 fragment
=escape_rfc3986(url_parsed
.fragment
)
2900 return compat_parse_qs(compat_urllib_parse_urlparse(url
).query
)
2903 def read_batch_urls(batch_fd
):
2905 if not isinstance(url
, compat_str
):
2906 url
= url
.decode('utf-8', 'replace')
2907 BOM_UTF8
= ('\xef\xbb\xbf', '\ufeff')
2908 for bom
in BOM_UTF8
:
2909 if url
.startswith(bom
):
2910 url
= url
[len(bom
):]
2912 if not url
or url
.startswith(('#', ';', ']')):
2914 # "#" cannot be stripped out since it is part of the URI
2915 # However, it can be safely stipped out if follwing a whitespace
2916 return re
.split(r
'\s#', url
, 1)[0].rstrip()
2918 with contextlib
.closing(batch_fd
) as fd
:
2919 return [url
for url
in map(fixup
, fd
) if url
]
2922 def urlencode_postdata(*args
, **kargs
):
2923 return compat_urllib_parse_urlencode(*args
, **kargs
).encode('ascii')
2926 def update_url_query(url
, query
):
2929 parsed_url
= compat_urlparse
.urlparse(url
)
2930 qs
= compat_parse_qs(parsed_url
.query
)
2932 return compat_urlparse
.urlunparse(parsed_url
._replace
(
2933 query
=compat_urllib_parse_urlencode(qs
, True)))
2936 def update_Request(req
, url
=None, data
=None, headers
={}, query={}
):
2937 req_headers
= req
.headers
.copy()
2938 req_headers
.update(headers
)
2939 req_data
= data
or req
.data
2940 req_url
= update_url_query(url
or req
.get_full_url(), query
)
2941 req_get_method
= req
.get_method()
2942 if req_get_method
== 'HEAD':
2943 req_type
= HEADRequest
2944 elif req_get_method
== 'PUT':
2945 req_type
= PUTRequest
2947 req_type
= compat_urllib_request
.Request
2949 req_url
, data
=req_data
, headers
=req_headers
,
2950 origin_req_host
=req
.origin_req_host
, unverifiable
=req
.unverifiable
)
2951 if hasattr(req
, 'timeout'):
2952 new_req
.timeout
= req
.timeout
2956 def _multipart_encode_impl(data
, boundary
):
2957 content_type
= 'multipart/form-data; boundary=%s' % boundary
2960 for k
, v
in data
.items():
2961 out
+= b
'--' + boundary
.encode('ascii') + b
'\r\n'
2962 if isinstance(k
, compat_str
):
2963 k
= k
.encode('utf-8')
2964 if isinstance(v
, compat_str
):
2965 v
= v
.encode('utf-8')
2966 # RFC 2047 requires non-ASCII field names to be encoded, while RFC 7578
2967 # suggests sending UTF-8 directly. Firefox sends UTF-8, too
2968 content
= b
'Content-Disposition: form-data; name="' + k
+ b
'"\r\n\r\n' + v
+ b
'\r\n'
2969 if boundary
.encode('ascii') in content
:
2970 raise ValueError('Boundary overlaps with data')
2973 out
+= b
'--' + boundary
.encode('ascii') + b
'--\r\n'
2975 return out
, content_type
2978 def multipart_encode(data
, boundary
=None):
2980 Encode a dict to RFC 7578-compliant form-data
2983 A dict where keys and values can be either Unicode or bytes-like
2986 If specified a Unicode object, it's used as the boundary. Otherwise
2987 a random boundary is generated.
2989 Reference: https://tools.ietf.org/html/rfc7578
2991 has_specified_boundary
= boundary
is not None
2994 if boundary
is None:
2995 boundary
= '---------------' + str(random
.randrange(0x0fffffff, 0xffffffff))
2998 out
, content_type
= _multipart_encode_impl(data
, boundary
)
3001 if has_specified_boundary
:
3005 return out
, content_type
3008 def dict_get(d
, key_or_keys
, default
=None, skip_false_values
=True):
3009 if isinstance(key_or_keys
, (list, tuple)):
3010 for key
in key_or_keys
:
3011 if key
not in d
or d
[key
] is None or skip_false_values
and not d
[key
]:
3015 return d
.get(key_or_keys
, default
)
3018 def try_get(src
, getter
, expected_type
=None):
3019 for get
in variadic(getter
):
3022 except (AttributeError, KeyError, TypeError, IndexError):
3025 if expected_type
is None or isinstance(v
, expected_type
):
3029 def merge_dicts(*dicts
):
3031 for a_dict
in dicts
:
3032 for k
, v
in a_dict
.items():
3036 or (isinstance(v
, compat_str
) and v
3037 and isinstance(merged
[k
], compat_str
)
3038 and not merged
[k
])):
3043 def encode_compat_str(string
, encoding
=preferredencoding(), errors
='strict'):
3044 return string
if isinstance(string
, compat_str
) else compat_str(string
, encoding
, errors
)
3056 TV_PARENTAL_GUIDELINES
= {
3066 def parse_age_limit(s
):
3068 return s
if 0 <= s
<= 21 else None
3069 if not isinstance(s
, compat_basestring
):
3071 m
= re
.match(r
'^(?P<age>\d{1,2})\+?$', s
)
3073 return int(m
.group('age'))
3076 return US_RATINGS
[s
]
3077 m
= re
.match(r
'^TV[_-]?(%s)$' % '|'.join(k
[3:] for k
in TV_PARENTAL_GUIDELINES
), s
)
3079 return TV_PARENTAL_GUIDELINES
['TV-' + m
.group(1)]
3083 def strip_jsonp(code
):
3086 (?:window\.)?(?P<func_name>[a-zA-Z0-9_.$]*)
3087 (?:\s*&&\s*(?P=func_name))?
3088 \s*\(\s*(?P<callback_data>.*)\);?
3089 \s*?(?://[^\n]*)*$''',
3090 r
'\g<callback_data>', code
)
3093 def js_to_json(code
, vars={}):
3094 # vars is a dict of var, val pairs to substitute
3095 COMMENT_RE
= r
'/\*(?:(?!\*/).)*?\*/|//[^\n]*\n'
3096 SKIP_RE
= r
'\s*(?:{comment})?\s*'.format(comment
=COMMENT_RE
)
3098 (r
'(?s)^(0[xX][0-9a-fA-F]+){skip}:?$'.format(skip
=SKIP_RE
), 16),
3099 (r
'(?s)^(0+[0-7]+){skip}:?$'.format(skip
=SKIP_RE
), 8),
3104 if v
in ('true', 'false', 'null'):
3106 elif v
in ('undefined', 'void 0'):
3108 elif v
.startswith('/*') or v
.startswith('//') or v
.startswith('!') or v
== ',':
3111 if v
[0] in ("'", '"'):
3112 v
= re
.sub(r
'(?s)\\.|"', lambda m
: {
3117 }.get(m
.group(0), m
.group(0)), v
[1:-1])
3119 for regex
, base
in INTEGER_TABLE
:
3120 im
= re
.match(regex
, v
)
3122 i
= int(im
.group(1), base
)
3123 return '"%d":' % i
if v
.endswith(':') else '%d' % i
3130 return re
.sub(r
'''(?sx)
3131 "(?:[^"\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^"\\]*"|
3132 '(?:[^'\\]*(?:\\\\|\\['"nurtbfx/\n]))*[^'\\]*'|
3133 {comment}|,(?={skip}[\]}}])|
3134 void\s0|(?:(?<![0-9])[eE]|[a-df-zA-DF-Z_$])[.a-zA-Z_$0-9]*|
3135 \b(?:0[xX][0-9a-fA-F]+|0+[0-7]+)(?:{skip}:)?|
3138 '''.format(comment
=COMMENT_RE
, skip
=SKIP_RE
), fix_kv
, code
)
3141 def qualities(quality_ids
):
3142 """ Get a numeric quality value out of a list of possible values """
3145 return quality_ids
.index(qid
)
3151 POSTPROCESS_WHEN
= {'pre_process', 'before_dl', 'after_move', 'post_process', 'after_video', 'playlist'}
3155 'default': '%(title)s [%(id)s].%(ext)s',
3156 'chapter': '%(title)s - %(section_number)03d %(section_title)s [%(id)s].%(ext)s',
3162 'description': 'description',
3163 'annotation': 'annotations.xml',
3164 'infojson': 'info.json',
3167 'pl_thumbnail': None,
3168 'pl_description': 'description',
3169 'pl_infojson': 'info.json',
3172 # As of [1] format syntax is:
3173 # %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
3174 # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
3175 STR_FORMAT_RE_TMPL
= r
'''(?x)
3176 (?<!%)(?P<prefix>(?:%%)*)
3178 (?P<has_key>\((?P<key>{0})\))?
3180 (?P<conversion>[#0\-+ ]+)?
3182 (?P<precision>\.\d+)?
3183 (?P<len_mod>[hlL])? # unused in python
3184 {1} # conversion type
3189 STR_FORMAT_TYPES
= 'diouxXeEfFgGcrs'
3192 def limit_length(s
, length
):
3193 """ Add ellipses to overly long strings """
3198 return s
[:length
- len(ELLIPSES
)] + ELLIPSES
3202 def version_tuple(v
):
3203 return tuple(int(e
) for e
in re
.split(r
'[-.]', v
))
3206 def is_outdated_version(version
, limit
, assume_new
=True):
3208 return not assume_new
3210 return version_tuple(version
) < version_tuple(limit
)
3212 return not assume_new
3215 def ytdl_is_updateable():
3216 """ Returns if yt-dlp can be updated with -U """
3218 from .update
import is_non_updateable
3220 return not is_non_updateable()
3223 def args_to_str(args
):
3224 # Get a short string representation for a subprocess command
3225 return ' '.join(compat_shlex_quote(a
) for a
in args
)
3228 def error_to_compat_str(err
):
3230 # On python 2 error byte string must be decoded with proper
3231 # encoding rather than ascii
3232 if sys
.version_info
[0] < 3:
3233 err_str
= err_str
.decode(preferredencoding())
3237 def mimetype2ext(mt
):
3241 mt
, _
, params
= mt
.partition(';')
3246 # Per RFC 3003, audio/mpeg can be .mp1, .mp2 or .mp3. Here use .mp3 as
3247 # it's the most popular one
3248 'audio/mpeg': 'mp3',
3249 'audio/x-wav': 'wav',
3251 'audio/wave': 'wav',
3254 ext
= FULL_MAP
.get(mt
)
3260 'smptett+xml': 'tt',
3264 'x-mp4-fragmented': 'mp4',
3265 'x-ms-sami': 'sami',
3268 'x-mpegurl': 'm3u8',
3269 'vnd.apple.mpegurl': 'm3u8',
3273 'vnd.ms-sstr+xml': 'ism',
3277 'filmstrip+json': 'fs',
3281 _
, _
, subtype
= mt
.rpartition('/')
3282 ext
= SUBTYPE_MAP
.get(subtype
.lower())
3293 _
, _
, suffix
= subtype
.partition('+')
3294 ext
= SUFFIX_MAP
.get(suffix
)
3298 return subtype
.replace('+', '.')
3301 def ext2mimetype(ext_or_url
):
3304 if '.' not in ext_or_url
:
3305 ext_or_url
= f
'file.{ext_or_url}'
3306 return mimetypes
.guess_type(ext_or_url
)[0]
3309 def parse_codecs(codecs_str
):
3310 # http://tools.ietf.org/html/rfc6381
3313 split_codecs
= list(filter(None, map(
3314 str.strip
, codecs_str
.strip().strip(',').split(','))))
3315 vcodec
, acodec
, tcodec
, hdr
= None, None, None, None
3316 for full_codec
in split_codecs
:
3317 parts
= full_codec
.split('.')
3318 codec
= parts
[0].replace('0', '')
3319 if codec
in ('avc1', 'avc2', 'avc3', 'avc4', 'vp9', 'vp8', 'hev1', 'hev2',
3320 'h263', 'h264', 'mp4v', 'hvc1', 'av1', 'theora', 'dvh1', 'dvhe'):
3322 vcodec
= '.'.join(parts
[:4]) if codec
in ('vp9', 'av1', 'hvc1') else full_codec
3323 if codec
in ('dvh1', 'dvhe'):
3325 elif codec
== 'av1' and len(parts
) > 3 and parts
[3] == '10':
3327 elif full_codec
.replace('0', '').startswith('vp9.2'):
3329 elif codec
in ('flac', 'mp4a', 'opus', 'vorbis', 'mp3', 'aac', 'ac-3', 'ec-3', 'eac3', 'dtsc', 'dtse', 'dtsh', 'dtsl'):
3332 elif codec
in ('stpp', 'wvtt',):
3336 write_string('WARNING: Unknown codec %s\n' % full_codec
, sys
.stderr
)
3337 if vcodec
or acodec
or tcodec
:
3339 'vcodec': vcodec
or 'none',
3340 'acodec': acodec
or 'none',
3341 'dynamic_range': hdr
,
3342 **({'tcodec': tcodec}
if tcodec
is not None else {}),
3344 elif len(split_codecs
) == 2:
3346 'vcodec': split_codecs
[0],
3347 'acodec': split_codecs
[1],
3352 def urlhandle_detect_ext(url_handle
):
3353 getheader
= url_handle
.headers
.get
3355 cd
= getheader('Content-Disposition')
3357 m
= re
.match(r
'attachment;\s*filename="(?P<filename>[^"]+)"', cd
)
3359 e
= determine_ext(m
.group('filename'), default_ext
=None)
3363 return mimetype2ext(getheader('Content-Type'))
3366 def encode_data_uri(data
, mime_type
):
3367 return 'data:%s;base64,%s' % (mime_type
, base64
.b64encode(data
).decode('ascii'))
3370 def age_restricted(content_limit
, age_limit
):
3371 """ Returns True iff the content should be blocked """
3373 if age_limit
is None: # No limit set
3375 if content_limit
is None:
3376 return False # Content available for everyone
3377 return age_limit
< content_limit
3380 def is_html(first_bytes
):
3381 """ Detect whether a file contains HTML by examining its first bytes. """
3384 (b
'\xef\xbb\xbf', 'utf-8'),
3385 (b
'\x00\x00\xfe\xff', 'utf-32-be'),
3386 (b
'\xff\xfe\x00\x00', 'utf-32-le'),
3387 (b
'\xff\xfe', 'utf-16-le'),
3388 (b
'\xfe\xff', 'utf-16-be'),
3390 for bom
, enc
in BOMS
:
3391 if first_bytes
.startswith(bom
):
3392 s
= first_bytes
[len(bom
):].decode(enc
, 'replace')
3395 s
= first_bytes
.decode('utf-8', 'replace')
3397 return re
.match(r
'^\s*<', s
)
3400 def determine_protocol(info_dict
):
3401 protocol
= info_dict
.get('protocol')
3402 if protocol
is not None:
3405 url
= sanitize_url(info_dict
['url'])
3406 if url
.startswith('rtmp'):
3408 elif url
.startswith('mms'):
3410 elif url
.startswith('rtsp'):
3413 ext
= determine_ext(url
)
3419 return compat_urllib_parse_urlparse(url
).scheme
3422 def render_table(header_row
, data
, delim
=False, extra_gap
=0, hide_empty
=False):
3423 """ Render a list of rows, each as a list of values.
3424 Text after a \t will be right aligned """
3426 return len(remove_terminal_sequences(string
).replace('\t', ''))
3428 def get_max_lens(table
):
3429 return [max(width(str(v
)) for v
in col
) for col
in zip(*table
)]
3431 def filter_using_list(row
, filterArray
):
3432 return [col
for take
, col
in itertools
.zip_longest(filterArray
, row
, fillvalue
=True) if take
]
3434 max_lens
= get_max_lens(data
) if hide_empty
else []
3435 header_row
= filter_using_list(header_row
, max_lens
)
3436 data
= [filter_using_list(row
, max_lens
) for row
in data
]
3438 table
= [header_row
] + data
3439 max_lens
= get_max_lens(table
)
3442 table
= [header_row
, [delim
* (ml
+ extra_gap
) for ml
in max_lens
]] + data
3443 table
[1][-1] = table
[1][-1][:-extra_gap
] # Remove extra_gap from end of delimiter
3445 for pos
, text
in enumerate(map(str, row
)):
3447 row
[pos
] = text
.replace('\t', ' ' * (max_lens
[pos
] - width(text
))) + ' ' * extra_gap
3449 row
[pos
] = text
+ ' ' * (max_lens
[pos
] - width(text
) + extra_gap
)
3450 ret
= '\n'.join(''.join(row
).rstrip() for row
in table
)
3454 def _match_one(filter_part
, dct
, incomplete
):
3455 # TODO: Generalize code with YoutubeDL._build_format_filter
3456 STRING_OPERATORS
= {
3457 '*=': operator
.contains
,
3458 '^=': lambda attr
, value
: attr
.startswith(value
),
3459 '$=': lambda attr
, value
: attr
.endswith(value
),
3460 '~=': lambda attr
, value
: re
.search(value
, attr
),
3462 COMPARISON_OPERATORS
= {
3464 '<=': operator
.le
, # "<=" must be defined above "<"
3471 operator_rex
= re
.compile(r
'''(?x)\s*
3473 \s*(?P<negation>!\s*)?(?P<op>%s)(?P<none_inclusive>\s*\?)?\s*
3475 (?P<quote>["\'])(?P
<quotedstrval
>.+?
)(?P
=quote
)|
3479 ''' % '|'.join(map(re.escape, COMPARISON_OPERATORS.keys())))
3480 m = operator_rex.search(filter_part)
3483 unnegated_op = COMPARISON_OPERATORS[m['op']]
3485 op = lambda attr, value: not unnegated_op(attr, value)
3488 comparison_value = m['quotedstrval'] or m['strval'] or m['intval']
3490 comparison_value = comparison_value.replace(r'\%s' % m['quote'], m['quote'])
3491 actual_value = dct.get(m['key'])
3492 numeric_comparison = None
3493 if isinstance(actual_value, compat_numeric_types):
3494 # If the original field is a string and matching comparisonvalue is
3495 # a number we should respect the origin of the original field
3496 # and process comparison value as a string (see
3497 # https://github.com/ytdl-org/youtube-dl/issues/11082)
3499 numeric_comparison = int(comparison_value)
3501 numeric_comparison = parse_filesize(comparison_value)
3502 if numeric_comparison is None:
3503 numeric_comparison = parse_filesize(f'{comparison_value}B')
3504 if numeric_comparison is None:
3505 numeric_comparison = parse_duration(comparison_value)
3506 if numeric_comparison is not None and m['op'] in STRING_OPERATORS:
3507 raise ValueError('Operator %s only supports string values!' % m['op'])
3508 if actual_value is None:
3509 return incomplete or m['none_inclusive']
3510 return op(actual_value, comparison_value if numeric_comparison is None else numeric_comparison)
3513 '': lambda v: (v is True) if isinstance(v, bool) else (v is not None),
3514 '!': lambda v: (v is False) if isinstance(v, bool) else (v is None),
3516 operator_rex = re.compile(r'''(?x
)\s
*
3517 (?P
<op
>%s)\s
*(?P
<key
>[a
-z_
]+)
3519 ''' % '|'.join(map(re.escape, UNARY_OPERATORS.keys())))
3520 m = operator_rex.search(filter_part)
3522 op = UNARY_OPERATORS[m.group('op')]
3523 actual_value = dct.get(m.group('key'))
3524 if incomplete and actual_value is None:
3526 return op(actual_value)
3528 raise ValueError('Invalid filter part %r' % filter_part)
3531 def match_str(filter_str, dct, incomplete=False):
3532 """ Filter a dictionary with a simple string syntax. Returns True (=passes filter) or false
3533 When incomplete, all conditions passes on missing fields
3536 _match_one(filter_part.replace(r'\&', '&'), dct, incomplete)
3537 for filter_part in re.split(r'(?<!\\)&', filter_str))
3540 def match_filter_func(filter_str):
3541 def _match_func(info_dict, *args, **kwargs):
3542 if match_str(filter_str, info_dict, *args, **kwargs):
3545 video_title = info_dict.get('title', info_dict.get('id', 'video'))
3546 return '%s does not pass filter %s, skipping ..' % (video_title, filter_str)
3550 def parse_dfxp_time_expr(time_expr):
3554 mobj = re.match(r'^(?P<time_offset>\d+(?:\.\d+)?)s?$', time_expr)
3556 return float(mobj.group('time_offset'))
3558 mobj = re.match(r'^(\d+):(\d\d):(\d\d(?:(?:\.|:)\d+)?)$', time_expr)
3560 return 3600 * int(mobj.group(1)) + 60 * int(mobj.group(2)) + float(mobj.group(3).replace(':', '.'))
3563 def srt_subtitles_timecode(seconds):
3564 return '%02d:%02d:%02d,%03d' % timetuple_from_msec(seconds * 1000)
3567 def ass_subtitles_timecode(seconds):
3568 time = timetuple_from_msec(seconds * 1000)
3569 return '%01d:%02d:%02d.%02d' % (*time[:-1], time.milliseconds / 10)
3572 def dfxp2srt(dfxp_data):
3574 @param dfxp_data A
bytes-like
object containing DFXP data
3575 @returns A
unicode object containing converted SRT data
3577 LEGACY_NAMESPACES = (
3578 (b'http://www.w3.org/ns/ttml', [
3579 b'http://www.w3.org/2004/11/ttaf1',
3580 b'http://www.w3.org/2006/04/ttaf1',
3581 b'http://www.w3.org/2006/10/ttaf1',
3583 (b'http://www.w3.org/ns/ttml#styling', [
3584 b'http://www.w3.org/ns/ttml#style',
3588 SUPPORTED_STYLING = [
3597 _x = functools.partial(xpath_with_ns, ns_map={
3598 'xml': 'http://www.w3.org/XML/1998/namespace',
3599 'ttml': 'http://www.w3.org/ns/ttml',
3600 'tts': 'http://www.w3.org/ns/ttml#styling',
3606 class TTMLPElementParser(object):
3608 _unclosed_elements = []
3609 _applied_styles = []
3611 def start(self, tag, attrib):
3612 if tag in (_x('ttml:br'), 'br'):
3615 unclosed_elements = []
3617 element_style_id = attrib.get('style')
3619 style.update(default_style)
3620 if element_style_id:
3621 style.update(styles.get(element_style_id, {}))
3622 for prop in SUPPORTED_STYLING:
3623 prop_val = attrib.get(_x('tts:' + prop))
3625 style[prop] = prop_val
3628 for k, v in sorted(style.items()):
3629 if self._applied_styles and self._applied_styles[-1].get(k) == v:
3632 font += ' color="%s"' % v
3633 elif k == 'fontSize':
3634 font += ' size="%s"' % v
3635 elif k == 'fontFamily':
3636 font += ' face="%s"' % v
3637 elif k == 'fontWeight' and v == 'bold':
3639 unclosed_elements.append('b')
3640 elif k == 'fontStyle' and v == 'italic':
3642 unclosed_elements.append('i')
3643 elif k == 'textDecoration' and v == 'underline':
3645 unclosed_elements.append('u')
3647 self._out += '<font' + font + '>'
3648 unclosed_elements.append('font')
3650 if self._applied_styles:
3651 applied_style.update(self._applied_styles[-1])
3652 applied_style.update(style)
3653 self._applied_styles.append(applied_style)
3654 self._unclosed_elements.append(unclosed_elements)
3657 if tag not in (_x('ttml:br'), 'br'):
3658 unclosed_elements = self._unclosed_elements.pop()
3659 for element in reversed(unclosed_elements):
3660 self._out += '</%s>' % element
3661 if unclosed_elements and self._applied_styles:
3662 self._applied_styles.pop()
3664 def data(self, data):
3668 return self._out.strip()
3670 def parse_node(node):
3671 target = TTMLPElementParser()
3672 parser = xml.etree.ElementTree.XMLParser(target=target)
3673 parser.feed(xml.etree.ElementTree.tostring(node))
3674 return parser.close()
3676 for k, v in LEGACY_NAMESPACES:
3678 dfxp_data = dfxp_data.replace(ns, k)
3680 dfxp = compat_etree_fromstring(dfxp_data)
3682 paras = dfxp.findall(_x('.//ttml:p')) or dfxp.findall('.//p')
3685 raise ValueError('Invalid dfxp/TTML subtitle')
3689 for style in dfxp.findall(_x('.//ttml:style')):
3690 style_id = style.get('id') or style.get(_x('xml:id'))
3693 parent_style_id = style.get('style')
3695 if parent_style_id not in styles:
3698 styles[style_id] = styles[parent_style_id].copy()
3699 for prop in SUPPORTED_STYLING:
3700 prop_val = style.get(_x('tts:' + prop))
3702 styles.setdefault(style_id, {})[prop] = prop_val
3708 for p in ('body', 'div'):
3709 ele = xpath_element(dfxp, [_x('.//ttml:' + p), './/' + p])
3712 style = styles.get(ele.get('style'))
3715 default_style.update(style)
3717 for para, index in zip(paras, itertools.count(1)):
3718 begin_time = parse_dfxp_time_expr(para.attrib.get('begin'))
3719 end_time = parse_dfxp_time_expr(para.attrib.get('end'))
3720 dur = parse_dfxp_time_expr(para.attrib.get('dur'))
3721 if begin_time is None:
3726 end_time = begin_time + dur
3727 out.append('%d\n%s --> %s\n%s\n\n' % (
3729 srt_subtitles_timecode(begin_time),
3730 srt_subtitles_timecode(end_time),
3736 def cli_option(params, command_option, param):
3737 param = params.get(param)
3739 param = compat_str(param)
3740 return [command_option, param] if param is not None else []
3743 def cli_bool_option(params, command_option, param, true_value='true', false_value='false', separator=None):
3744 param = params.get(param)
3747 assert isinstance(param, bool)
3749 return [command_option + separator + (true_value if param else false_value)]
3750 return [command_option, true_value if param else false_value]
3753 def cli_valueless_option(params, command_option, param, expected_value=True):
3754 param = params.get(param)
3755 return [command_option] if param == expected_value else []
3758 def cli_configuration_args(argdict, keys, default=[], use_compat=True):
3759 if isinstance(argdict, (list, tuple)): # for backward compatibility
3766 assert isinstance(argdict, dict)
3768 assert isinstance(keys, (list, tuple))
3769 for key_list in keys:
3770 arg_list = list(filter(
3771 lambda x: x is not None,
3772 [argdict.get(key.lower()) for key in variadic(key_list)]))
3774 return [arg for args in arg_list for arg in args]
3778 def _configuration_args(main_key, argdict, exe, keys=None, default=[], use_compat=True):
3779 main_key, exe = main_key.lower(), exe.lower()
3780 root_key = exe if main_key == exe else f'{main_key}+{exe}'
3781 keys = [f'{root_key}{k}' for k in (keys or [''])]
3782 if root_key in keys:
3784 keys.append((main_key, exe))
3785 keys.append('default')
3788 return cli_configuration_args(argdict, keys, default, use_compat)
3791 class ISO639Utils(object):
3792 # See http://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
3851 'iw': 'heb', # Replaced by he in 1989 revision
3861 'in': 'ind', # Replaced by id in 1989 revision
3976 'ji': 'yid', # Replaced by yi in 1989 revision
3984 def short2long(cls, code):
3985 """Convert language code from ISO 639-1 to ISO 639-2/T"""
3986 return cls._lang_map.get(code[:2])
3989 def long2short(cls, code):
3990 """Convert language code from ISO 639-2/T to ISO 639-1"""
3991 for short_name, long_name in cls._lang_map.items():
3992 if long_name == code:
3996 class ISO3166Utils(object):
3997 # From http://data.okfn.org/data/core/country-list
3999 'AF': 'Afghanistan',
4000 'AX': 'Åland Islands',
4003 'AS': 'American Samoa',
4008 'AG': 'Antigua and Barbuda',
4025 'BO': 'Bolivia, Plurinational State of',
4026 'BQ': 'Bonaire, Sint Eustatius and Saba',
4027 'BA': 'Bosnia and Herzegovina',
4029 'BV': 'Bouvet Island',
4031 'IO': 'British Indian Ocean Territory',
4032 'BN': 'Brunei Darussalam',
4034 'BF': 'Burkina Faso',
4040 'KY': 'Cayman Islands',
4041 'CF': 'Central African Republic',
4045 'CX': 'Christmas Island',
4046 'CC': 'Cocos (Keeling) Islands',
4050 'CD': 'Congo, the Democratic Republic of the',
4051 'CK': 'Cook Islands',
4053 'CI': 'Côte d\'Ivoire',
4058 'CZ': 'Czech Republic',
4062 'DO': 'Dominican Republic',
4065 'SV': 'El Salvador',
4066 'GQ': 'Equatorial Guinea',
4070 'FK': 'Falkland Islands (Malvinas)',
4071 'FO': 'Faroe Islands',
4075 'GF': 'French Guiana',
4076 'PF': 'French Polynesia',
4077 'TF': 'French Southern Territories',
4092 'GW': 'Guinea-Bissau',
4095 'HM': 'Heard Island and McDonald Islands',
4096 'VA': 'Holy See (Vatican City State)',
4103 'IR': 'Iran, Islamic Republic of',
4106 'IM': 'Isle of Man',
4116 'KP': 'Korea, Democratic People\'s Republic of',
4117 'KR': 'Korea, Republic of',
4120 'LA': 'Lao People\'s Democratic Republic',
4126 'LI': 'Liechtenstein',
4130 'MK': 'Macedonia, the Former Yugoslav Republic of',
4137 'MH': 'Marshall Islands',
4143 'FM': 'Micronesia, Federated States of',
4144 'MD': 'Moldova, Republic of',
4155 'NL': 'Netherlands',
4156 'NC': 'New Caledonia',
4157 'NZ': 'New Zealand',
4162 'NF': 'Norfolk Island',
4163 'MP': 'Northern Mariana Islands',
4168 'PS': 'Palestine, State of',
4170 'PG': 'Papua New Guinea',
4173 'PH': 'Philippines',
4177 'PR': 'Puerto Rico',
4181 'RU': 'Russian Federation',
4183 'BL': 'Saint Barthélemy',
4184 'SH': 'Saint Helena, Ascension and Tristan da Cunha',
4185 'KN': 'Saint Kitts and Nevis',
4186 'LC': 'Saint Lucia',
4187 'MF': 'Saint Martin (French part)',
4188 'PM': 'Saint Pierre and Miquelon',
4189 'VC': 'Saint Vincent and the Grenadines',
4192 'ST': 'Sao Tome and Principe',
4193 'SA': 'Saudi Arabia',
4197 'SL': 'Sierra Leone',
4199 'SX': 'Sint Maarten (Dutch part)',
4202 'SB': 'Solomon Islands',
4204 'ZA': 'South Africa',
4205 'GS': 'South Georgia and the South Sandwich Islands',
4206 'SS': 'South Sudan',
4211 'SJ': 'Svalbard and Jan Mayen',
4214 'CH': 'Switzerland',
4215 'SY': 'Syrian Arab Republic',
4216 'TW': 'Taiwan, Province of China',
4218 'TZ': 'Tanzania, United Republic of',
4220 'TL': 'Timor-Leste',
4224 'TT': 'Trinidad and Tobago',
4227 'TM': 'Turkmenistan',
4228 'TC': 'Turks and Caicos Islands',
4232 'AE': 'United Arab Emirates',
4233 'GB': 'United Kingdom',
4234 'US': 'United States',
4235 'UM': 'United States Minor Outlying Islands',
4239 'VE': 'Venezuela, Bolivarian Republic of',
4241 'VG': 'Virgin Islands, British',
4242 'VI': 'Virgin Islands, U.S.',
4243 'WF': 'Wallis and Futuna',
4244 'EH': 'Western Sahara',
4251 def short2full(cls, code):
4252 """Convert an ISO 3166-2 country code to the corresponding full name"""
4253 return cls._country_map.get(code.upper())
4256 class GeoUtils(object):
4257 # Major IPv4 address blocks per country
4259 'AD': '46.172.224.0/19',
4260 'AE': '94.200.0.0/13',
4261 'AF': '149.54.0.0/17',
4262 'AG': '209.59.64.0/18',
4263 'AI': '204.14.248.0/21',
4264 'AL': '46.99.0.0/16',
4265 'AM': '46.70.0.0/15',
4266 'AO': '105.168.0.0/13',
4267 'AP': '182.50.184.0/21',
4268 'AQ': '23.154.160.0/24',
4269 'AR': '181.0.0.0/12',
4270 'AS': '202.70.112.0/20',
4271 'AT': '77.116.0.0/14',
4272 'AU': '1.128.0.0/11',
4273 'AW': '181.41.0.0/18',
4274 'AX': '185.217.4.0/22',
4275 'AZ': '5.197.0.0/16',
4276 'BA': '31.176.128.0/17',
4277 'BB': '65.48.128.0/17',
4278 'BD': '114.130.0.0/16',
4280 'BF': '102.178.0.0/15',
4281 'BG': '95.42.0.0/15',
4282 'BH': '37.131.0.0/17',
4283 'BI': '154.117.192.0/18',
4284 'BJ': '137.255.0.0/16',
4285 'BL': '185.212.72.0/23',
4286 'BM': '196.12.64.0/18',
4287 'BN': '156.31.0.0/16',
4288 'BO': '161.56.0.0/16',
4289 'BQ': '161.0.80.0/20',
4290 'BR': '191.128.0.0/12',
4291 'BS': '24.51.64.0/18',
4292 'BT': '119.2.96.0/19',
4293 'BW': '168.167.0.0/16',
4294 'BY': '178.120.0.0/13',
4295 'BZ': '179.42.192.0/18',
4296 'CA': '99.224.0.0/11',
4297 'CD': '41.243.0.0/16',
4298 'CF': '197.242.176.0/21',
4299 'CG': '160.113.0.0/16',
4300 'CH': '85.0.0.0/13',
4301 'CI': '102.136.0.0/14',
4302 'CK': '202.65.32.0/19',
4303 'CL': '152.172.0.0/14',
4304 'CM': '102.244.0.0/14',
4305 'CN': '36.128.0.0/10',
4306 'CO': '181.240.0.0/12',
4307 'CR': '201.192.0.0/12',
4308 'CU': '152.206.0.0/15',
4309 'CV': '165.90.96.0/19',
4310 'CW': '190.88.128.0/17',
4311 'CY': '31.153.0.0/16',
4312 'CZ': '88.100.0.0/14',
4314 'DJ': '197.241.0.0/17',
4315 'DK': '87.48.0.0/12',
4316 'DM': '192.243.48.0/20',
4317 'DO': '152.166.0.0/15',
4318 'DZ': '41.96.0.0/12',
4319 'EC': '186.68.0.0/15',
4320 'EE': '90.190.0.0/15',
4321 'EG': '156.160.0.0/11',
4322 'ER': '196.200.96.0/20',
4323 'ES': '88.0.0.0/11',
4324 'ET': '196.188.0.0/14',
4325 'EU': '2.16.0.0/13',
4326 'FI': '91.152.0.0/13',
4327 'FJ': '144.120.0.0/16',
4328 'FK': '80.73.208.0/21',
4329 'FM': '119.252.112.0/20',
4330 'FO': '88.85.32.0/19',
4332 'GA': '41.158.0.0/15',
4334 'GD': '74.122.88.0/21',
4335 'GE': '31.146.0.0/16',
4336 'GF': '161.22.64.0/18',
4337 'GG': '62.68.160.0/19',
4338 'GH': '154.160.0.0/12',
4339 'GI': '95.164.0.0/16',
4340 'GL': '88.83.0.0/19',
4341 'GM': '160.182.0.0/15',
4342 'GN': '197.149.192.0/18',
4343 'GP': '104.250.0.0/19',
4344 'GQ': '105.235.224.0/20',
4345 'GR': '94.64.0.0/13',
4346 'GT': '168.234.0.0/16',
4347 'GU': '168.123.0.0/16',
4348 'GW': '197.214.80.0/20',
4349 'GY': '181.41.64.0/18',
4350 'HK': '113.252.0.0/14',
4351 'HN': '181.210.0.0/16',
4352 'HR': '93.136.0.0/13',
4353 'HT': '148.102.128.0/17',
4354 'HU': '84.0.0.0/14',
4355 'ID': '39.192.0.0/10',
4356 'IE': '87.32.0.0/12',
4357 'IL': '79.176.0.0/13',
4358 'IM': '5.62.80.0/20',
4359 'IN': '117.192.0.0/10',
4360 'IO': '203.83.48.0/21',
4361 'IQ': '37.236.0.0/14',
4362 'IR': '2.176.0.0/12',
4363 'IS': '82.221.0.0/16',
4364 'IT': '79.0.0.0/10',
4365 'JE': '87.244.64.0/18',
4366 'JM': '72.27.0.0/17',
4367 'JO': '176.29.0.0/16',
4368 'JP': '133.0.0.0/8',
4369 'KE': '105.48.0.0/12',
4370 'KG': '158.181.128.0/17',
4371 'KH': '36.37.128.0/17',
4372 'KI': '103.25.140.0/22',
4373 'KM': '197.255.224.0/20',
4374 'KN': '198.167.192.0/19',
4375 'KP': '175.45.176.0/22',
4376 'KR': '175.192.0.0/10',
4377 'KW': '37.36.0.0/14',
4378 'KY': '64.96.0.0/15',
4379 'KZ': '2.72.0.0/13',
4380 'LA': '115.84.64.0/18',
4381 'LB': '178.135.0.0/16',
4382 'LC': '24.92.144.0/20',
4383 'LI': '82.117.0.0/19',
4384 'LK': '112.134.0.0/15',
4385 'LR': '102.183.0.0/16',
4386 'LS': '129.232.0.0/17',
4387 'LT': '78.56.0.0/13',
4388 'LU': '188.42.0.0/16',
4389 'LV': '46.109.0.0/16',
4390 'LY': '41.252.0.0/14',
4391 'MA': '105.128.0.0/11',
4392 'MC': '88.209.64.0/18',
4393 'MD': '37.246.0.0/16',
4394 'ME': '178.175.0.0/17',
4395 'MF': '74.112.232.0/21',
4396 'MG': '154.126.0.0/17',
4397 'MH': '117.103.88.0/21',
4398 'MK': '77.28.0.0/15',
4399 'ML': '154.118.128.0/18',
4400 'MM': '37.111.0.0/17',
4401 'MN': '49.0.128.0/17',
4402 'MO': '60.246.0.0/16',
4403 'MP': '202.88.64.0/20',
4404 'MQ': '109.203.224.0/19',
4405 'MR': '41.188.64.0/18',
4406 'MS': '208.90.112.0/22',
4407 'MT': '46.11.0.0/16',
4408 'MU': '105.16.0.0/12',
4409 'MV': '27.114.128.0/18',
4410 'MW': '102.70.0.0/15',
4411 'MX': '187.192.0.0/11',
4412 'MY': '175.136.0.0/13',
4413 'MZ': '197.218.0.0/15',
4414 'NA': '41.182.0.0/16',
4415 'NC': '101.101.0.0/18',
4416 'NE': '197.214.0.0/18',
4417 'NF': '203.17.240.0/22',
4418 'NG': '105.112.0.0/12',
4419 'NI': '186.76.0.0/15',
4420 'NL': '145.96.0.0/11',
4421 'NO': '84.208.0.0/13',
4422 'NP': '36.252.0.0/15',
4423 'NR': '203.98.224.0/19',
4424 'NU': '49.156.48.0/22',
4425 'NZ': '49.224.0.0/14',
4426 'OM': '5.36.0.0/15',
4427 'PA': '186.72.0.0/15',
4428 'PE': '186.160.0.0/14',
4429 'PF': '123.50.64.0/18',
4430 'PG': '124.240.192.0/19',
4431 'PH': '49.144.0.0/13',
4432 'PK': '39.32.0.0/11',
4433 'PL': '83.0.0.0/11',
4434 'PM': '70.36.0.0/20',
4435 'PR': '66.50.0.0/16',
4436 'PS': '188.161.0.0/16',
4437 'PT': '85.240.0.0/13',
4438 'PW': '202.124.224.0/20',
4439 'PY': '181.120.0.0/14',
4440 'QA': '37.210.0.0/15',
4441 'RE': '102.35.0.0/16',
4442 'RO': '79.112.0.0/13',
4443 'RS': '93.86.0.0/15',
4444 'RU': '5.136.0.0/13',
4445 'RW': '41.186.0.0/16',
4446 'SA': '188.48.0.0/13',
4447 'SB': '202.1.160.0/19',
4448 'SC': '154.192.0.0/11',
4449 'SD': '102.120.0.0/13',
4450 'SE': '78.64.0.0/12',
4451 'SG': '8.128.0.0/10',
4452 'SI': '188.196.0.0/14',
4453 'SK': '78.98.0.0/15',
4454 'SL': '102.143.0.0/17',
4455 'SM': '89.186.32.0/19',
4456 'SN': '41.82.0.0/15',
4457 'SO': '154.115.192.0/18',
4458 'SR': '186.179.128.0/17',
4459 'SS': '105.235.208.0/21',
4460 'ST': '197.159.160.0/19',
4461 'SV': '168.243.0.0/16',
4462 'SX': '190.102.0.0/20',
4464 'SZ': '41.84.224.0/19',
4465 'TC': '65.255.48.0/20',
4466 'TD': '154.68.128.0/19',
4467 'TG': '196.168.0.0/14',
4468 'TH': '171.96.0.0/13',
4469 'TJ': '85.9.128.0/18',
4470 'TK': '27.96.24.0/21',
4471 'TL': '180.189.160.0/20',
4472 'TM': '95.85.96.0/19',
4473 'TN': '197.0.0.0/11',
4474 'TO': '175.176.144.0/21',
4475 'TR': '78.160.0.0/11',
4476 'TT': '186.44.0.0/15',
4477 'TV': '202.2.96.0/19',
4478 'TW': '120.96.0.0/11',
4479 'TZ': '156.156.0.0/14',
4480 'UA': '37.52.0.0/14',
4481 'UG': '102.80.0.0/13',
4483 'UY': '167.56.0.0/13',
4484 'UZ': '84.54.64.0/18',
4485 'VA': '212.77.0.0/19',
4486 'VC': '207.191.240.0/21',
4487 'VE': '186.88.0.0/13',
4488 'VG': '66.81.192.0/20',
4489 'VI': '146.226.0.0/16',
4490 'VN': '14.160.0.0/11',
4491 'VU': '202.80.32.0/20',
4492 'WF': '117.20.32.0/21',
4493 'WS': '202.4.32.0/19',
4494 'YE': '134.35.0.0/16',
4495 'YT': '41.242.116.0/22',
4496 'ZA': '41.0.0.0/11',
4497 'ZM': '102.144.0.0/13',
4498 'ZW': '102.177.192.0/18',
4502 def random_ipv4(cls, code_or_block):
4503 if len(code_or_block) == 2:
4504 block = cls._country_ip_map.get(code_or_block.upper())
4508 block = code_or_block
4509 addr, preflen = block.split('/')
4510 addr_min = compat_struct_unpack('!L', socket.inet_aton(addr))[0]
4511 addr_max = addr_min | (0xffffffff >> int(preflen))
4512 return compat_str(socket.inet_ntoa(
4513 compat_struct_pack('!L', random.randint(addr_min, addr_max))))
4516 class PerRequestProxyHandler(compat_urllib_request.ProxyHandler):
4517 def __init__(self, proxies=None):
4518 # Set default handlers
4519 for type in ('http', 'https'):
4520 setattr(self, '%s_open' % type,
4521 lambda r, proxy='__noproxy__', type=type, meth=self.proxy_open:
4522 meth(r, proxy, type))
4523 compat_urllib_request.ProxyHandler.__init__(self, proxies)
4525 def proxy_open(self, req, proxy, type):
4526 req_proxy = req.headers.get('Ytdl-request-proxy')
4527 if req_proxy is not None:
4529 del req.headers['Ytdl-request-proxy']
4531 if proxy == '__noproxy__':
4532 return None # No Proxy
4533 if compat_urlparse.urlparse(proxy).scheme.lower() in ('socks', 'socks4', 'socks4a', 'socks5'):
4534 req.add_header('Ytdl-socks-proxy', proxy)
4535 # yt-dlp's http/https handlers do wrapping the socket with socks
4537 return compat_urllib_request.ProxyHandler.proxy_open(
4538 self, req, proxy, type)
4541 # Both long_to_bytes and bytes_to_long are adapted from PyCrypto, which is
4542 # released into Public Domain
4543 # https://github.com/dlitz/pycrypto/blob/master/lib/Crypto/Util/number.py#L387
4545 def long_to_bytes(n, blocksize=0):
4546 """long_to_bytes(n:long, blocksize:int) : string
4547 Convert a long integer to a byte string.
4549 If optional blocksize is given and greater than zero, pad the front of the
4550 byte string with binary zeros so that the length is a multiple of
4553 # after much testing, this algorithm was deemed to be the fastest
4557 s = compat_struct_pack('>I', n & 0xffffffff) + s
4559 # strip off leading zeros
4560 for i in range(len(s)):
4561 if s[i] != b'\000'[0]:
4564 # only happens when n == 0
4568 # add back some pad bytes. this could be done more efficiently w.r.t. the
4569 # de-padding being done above, but sigh...
4570 if blocksize > 0 and len(s) % blocksize:
4571 s = (blocksize - len(s) % blocksize) * b'\000' + s
4575 def bytes_to_long(s):
4576 """bytes_to_long(string) : long
4577 Convert a byte string to a long integer.
4579 This is (essentially) the inverse of long_to_bytes().
4584 extra = (4 - length % 4)
4585 s = b'\000' * extra + s
4586 length = length + extra
4587 for i in range(0, length, 4):
4588 acc = (acc << 32) + compat_struct_unpack('>I', s[i:i + 4])[0]
4592 def ohdave_rsa_encrypt(data, exponent, modulus):
4594 Implement OHDave
's RSA algorithm. See http://www.ohdave.com/rsa/
4597 data: data to encrypt, bytes-like object
4598 exponent, modulus: parameter e and N of RSA algorithm, both integer
4599 Output: hex string of encrypted data
4601 Limitation: supports one block encryption only
4604 payload = int(binascii.hexlify(data[::-1]), 16)
4605 encrypted = pow(payload, exponent, modulus)
4606 return '%x' % encrypted
4609 def pkcs1pad(data, length):
4611 Padding input data with PKCS#1 scheme
4613 @param {int[]} data input data
4614 @param {int} length target length
4615 @returns {int[]} padded data
4617 if len(data) > length - 11:
4618 raise ValueError('Input data too
long for PKCS
#1 padding')
4620 pseudo_random
= [random
.randint(0, 254) for _
in range(length
- len(data
) - 3)]
4621 return [0, 2] + pseudo_random
+ [0] + data
4624 def encode_base_n(num
, n
, table
=None):
4625 FULL_TABLE
= '0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ'
4627 table
= FULL_TABLE
[:n
]
4630 raise ValueError('base %d exceeds table length %d' % (n
, len(table
)))
4637 ret
= table
[num
% n
] + ret
4642 def decode_packed_codes(code
):
4643 mobj
= re
.search(PACKED_CODES_RE
, code
)
4644 obfuscated_code
, base
, count
, symbols
= mobj
.groups()
4647 symbols
= symbols
.split('|')
4652 base_n_count
= encode_base_n(count
, base
)
4653 symbol_table
[base_n_count
] = symbols
[count
] or base_n_count
4656 r
'\b(\w+)\b', lambda mobj
: symbol_table
[mobj
.group(0)],
4660 def caesar(s
, alphabet
, shift
):
4665 alphabet
[(alphabet
.index(c
) + shift
) % l
] if c
in alphabet
else c
4670 return caesar(s
, r
'''!"#$%&'()*+,-./0123456789:;<=>?@ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]^_`abcdefghijklmnopqrstuvwxyz{|}~''', 47)
4673 def parse_m3u8_attributes(attrib
):
4675 for (key
, val
) in re
.findall(r
'(?P<key>[A-Z0-9-]+)=(?P<val>"[^"]+"|[^",]+)(?:,|$)', attrib
):
4676 if val
.startswith('"'):
4682 def urshift(val
, n
):
4683 return val
>> n
if val
>= 0 else (val
+ 0x100000000) >> n
4686 # Based on png2str() written by @gdkchan and improved by @yokrysty
4687 # Originally posted at https://github.com/ytdl-org/youtube-dl/issues/9706
4688 def decode_png(png_data
):
4689 # Reference: https://www.w3.org/TR/PNG/
4690 header
= png_data
[8:]
4692 if png_data
[:8] != b
'\x89PNG\x0d\x0a\x1a\x0a' or header
[4:8] != b
'IHDR':
4693 raise IOError('Not a valid PNG file.')
4695 int_map
= {1: '>B', 2: '>H', 4: '>I'}
4696 unpack_integer
= lambda x
: compat_struct_unpack(int_map
[len(x
)], x
)[0]
4701 length
= unpack_integer(header
[:4])
4704 chunk_type
= header
[:4]
4707 chunk_data
= header
[:length
]
4708 header
= header
[length
:]
4710 header
= header
[4:] # Skip CRC
4718 ihdr
= chunks
[0]['data']
4720 width
= unpack_integer(ihdr
[:4])
4721 height
= unpack_integer(ihdr
[4:8])
4725 for chunk
in chunks
:
4726 if chunk
['type'] == b
'IDAT':
4727 idat
+= chunk
['data']
4730 raise IOError('Unable to read PNG data.')
4732 decompressed_data
= bytearray(zlib
.decompress(idat
))
4737 def _get_pixel(idx
):
4742 for y
in range(height
):
4743 basePos
= y
* (1 + stride
)
4744 filter_type
= decompressed_data
[basePos
]
4748 pixels
.append(current_row
)
4750 for x
in range(stride
):
4751 color
= decompressed_data
[1 + basePos
+ x
]
4752 basex
= y
* stride
+ x
4757 left
= _get_pixel(basex
- 3)
4759 up
= _get_pixel(basex
- stride
)
4761 if filter_type
== 1: # Sub
4762 color
= (color
+ left
) & 0xff
4763 elif filter_type
== 2: # Up
4764 color
= (color
+ up
) & 0xff
4765 elif filter_type
== 3: # Average
4766 color
= (color
+ ((left
+ up
) >> 1)) & 0xff
4767 elif filter_type
== 4: # Paeth
4773 c
= _get_pixel(basex
- stride
- 3)
4781 if pa
<= pb
and pa
<= pc
:
4782 color
= (color
+ a
) & 0xff
4784 color
= (color
+ b
) & 0xff
4786 color
= (color
+ c
) & 0xff
4788 current_row
.append(color
)
4790 return width
, height
, pixels
4793 def write_xattr(path
, key
, value
):
4794 # This mess below finds the best xattr tool for the job
4796 # try the pyxattr module...
4799 if hasattr(xattr
, 'set'): # pyxattr
4800 # Unicode arguments are not supported in python-pyxattr until
4802 # See https://github.com/ytdl-org/youtube-dl/issues/5498
4803 pyxattr_required_version
= '0.5.0'
4804 if version_tuple(xattr
.__version
__) < version_tuple(pyxattr_required_version
):
4805 # TODO: fallback to CLI tools
4806 raise XAttrUnavailableError(
4807 'python-pyxattr is detected but is too old. '
4808 'yt-dlp requires %s or above while your version is %s. '
4809 'Falling back to other xattr implementations' % (
4810 pyxattr_required_version
, xattr
.__version
__))
4812 setxattr
= xattr
.set
4814 setxattr
= xattr
.setxattr
4817 setxattr(path
, key
, value
)
4818 except EnvironmentError as e
:
4819 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4822 if compat_os_name
== 'nt':
4823 # Write xattrs to NTFS Alternate Data Streams:
4824 # http://en.wikipedia.org/wiki/NTFS#Alternate_data_streams_.28ADS.29
4825 assert ':' not in key
4826 assert os
.path
.exists(path
)
4828 ads_fn
= path
+ ':' + key
4830 with open(ads_fn
, 'wb') as f
:
4832 except EnvironmentError as e
:
4833 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4835 user_has_setfattr
= check_executable('setfattr', ['--version'])
4836 user_has_xattr
= check_executable('xattr', ['-h'])
4838 if user_has_setfattr
or user_has_xattr
:
4840 value
= value
.decode('utf-8')
4841 if user_has_setfattr
:
4842 executable
= 'setfattr'
4843 opts
= ['-n', key
, '-v', value
]
4844 elif user_has_xattr
:
4845 executable
= 'xattr'
4846 opts
= ['-w', key
, value
]
4848 cmd
= ([encodeFilename(executable
, True)]
4849 + [encodeArgument(o
) for o
in opts
]
4850 + [encodeFilename(path
, True)])
4854 cmd
, stdout
=subprocess
.PIPE
, stderr
=subprocess
.PIPE
, stdin
=subprocess
.PIPE
)
4855 except EnvironmentError as e
:
4856 raise XAttrMetadataError(e
.errno
, e
.strerror
)
4857 stdout
, stderr
= p
.communicate_or_kill()
4858 stderr
= stderr
.decode('utf-8', 'replace')
4859 if p
.returncode
!= 0:
4860 raise XAttrMetadataError(p
.returncode
, stderr
)
4863 # On Unix, and can't find pyxattr, setfattr, or xattr.
4864 if sys
.platform
.startswith('linux'):
4865 raise XAttrUnavailableError(
4866 "Couldn't find a tool to set the xattrs. "
4867 "Install either the python 'pyxattr' or 'xattr' "
4868 "modules, or the GNU 'attr' package "
4869 "(which contains the 'setfattr' tool).")
4871 raise XAttrUnavailableError(
4872 "Couldn't find a tool to set the xattrs. "
4873 "Install either the python 'xattr' module, "
4874 "or the 'xattr' binary.")
4877 def random_birthday(year_field
, month_field
, day_field
):
4878 start_date
= datetime
.date(1950, 1, 1)
4879 end_date
= datetime
.date(1995, 12, 31)
4880 offset
= random
.randint(0, (end_date
- start_date
).days
)
4881 random_date
= start_date
+ datetime
.timedelta(offset
)
4883 year_field
: str(random_date
.year
),
4884 month_field
: str(random_date
.month
),
4885 day_field
: str(random_date
.day
),
4889 # Templates for internet shortcut files, which are plain text files.
4890 DOT_URL_LINK_TEMPLATE
= '''
4895 DOT_WEBLOC_LINK_TEMPLATE
= '''
4896 <?xml version="1.0" encoding="UTF-8"?>
4897 <!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
4898 <plist version="1.0">
4901 \t<string>%(url)s</string>
4906 DOT_DESKTOP_LINK_TEMPLATE
= '''
4916 'url': DOT_URL_LINK_TEMPLATE
,
4917 'desktop': DOT_DESKTOP_LINK_TEMPLATE
,
4918 'webloc': DOT_WEBLOC_LINK_TEMPLATE
,
4922 def iri_to_uri(iri
):
4924 Converts an IRI (Internationalized Resource Identifier, allowing Unicode characters) to a URI (Uniform Resource Identifier, ASCII-only).
4926 The function doesn't add an additional layer of escaping; e.g., it doesn't escape `%3C` as `%253C`. Instead, it percent-escapes characters with an underlying UTF-8 encoding *besides* those already escaped, leaving the URI intact.
4929 iri_parts
= compat_urllib_parse_urlparse(iri
)
4931 if '[' in iri_parts
.netloc
:
4932 raise ValueError('IPv6 URIs are not, yet, supported.')
4933 # Querying `.netloc`, when there's only one bracket, also raises a ValueError.
4935 # The `safe` argument values, that the following code uses, contain the characters that should not be percent-encoded. Everything else but letters, digits and '_.-' will be percent-encoded with an underlying UTF-8 encoding. Everything already percent-encoded will be left as is.
4938 if iri_parts
.username
:
4939 net_location
+= compat_urllib_parse_quote(iri_parts
.username
, safe
=r
"!$%&'()*+,~")
4940 if iri_parts
.password
is not None:
4941 net_location
+= ':' + compat_urllib_parse_quote(iri_parts
.password
, safe
=r
"!$%&'()*+,~")
4944 net_location
+= iri_parts
.hostname
.encode('idna').decode('utf-8') # Punycode for Unicode hostnames.
4945 # The 'idna' encoding produces ASCII text.
4946 if iri_parts
.port
is not None and iri_parts
.port
!= 80:
4947 net_location
+= ':' + str(iri_parts
.port
)
4949 return compat_urllib_parse_urlunparse(
4953 compat_urllib_parse_quote_plus(iri_parts
.path
, safe
=r
"!$%&'()*+,/:;=@|~"),
4955 # Unsure about the `safe` argument, since this is a legacy way of handling parameters.
4956 compat_urllib_parse_quote_plus(iri_parts
.params
, safe
=r
"!$%&'()*+,/:;=@|~"),
4958 # Not totally sure about the `safe` argument, since the source does not explicitly mention the query URI component.
4959 compat_urllib_parse_quote_plus(iri_parts
.query
, safe
=r
"!$%&'()*+,/:;=?@{|}~"),
4961 compat_urllib_parse_quote_plus(iri_parts
.fragment
, safe
=r
"!#$%&'()*+,/:;=?@{|}~")))
4963 # Source for `safe` arguments: https://url.spec.whatwg.org/#percent-encoded-bytes.
4966 def to_high_limit_path(path
):
4967 if sys
.platform
in ['win32', 'cygwin']:
4968 # Work around MAX_PATH limitation on Windows. The maximum allowed length for the individual path segments may still be quite limited.
4969 return r
'\\?\ '.rstrip() + os
.path
.abspath(path
)
4974 def format_field(obj
, field
=None, template
='%s', ignore
=(None, ''), default
='', func
=None):
4975 val
= traverse_obj(obj
, *variadic(field
))
4978 return template
% (func(val
) if func
else val
)
4981 def clean_podcast_url(url
):
4982 return re
.sub(r
'''(?x)
4986 media\.blubrry\.com| # https://create.blubrry.com/resources/podcast-media-download-statistics/getting-started/
4989 (?:dts|www)\.podtrac\.com/(?:pts/)?redirect\.[0-9a-z]{3,4}| # http://analytics.podtrac.com/how-to-measure
4992 cn\.co| # https://podcorn.com/analytics-prefix/
4993 st\.fm # https://podsights.com/docs/
4998 _HEX_TABLE
= '0123456789abcdef'
5001 def random_uuidv4():
5002 return re
.sub(r
'[xy]', lambda x
: _HEX_TABLE
[random
.randint(0, 15)], 'xxxxxxxx-xxxx-4xxx-yxxx-xxxxxxxxxxxx')
5005 def make_dir(path
, to_screen
=None):
5007 dn
= os
.path
.dirname(path
)
5008 if dn
and not os
.path
.exists(dn
):
5011 except (OSError, IOError) as err
:
5012 if callable(to_screen
) is not None:
5013 to_screen('unable to create directory ' + error_to_compat_str(err
))
5017 def get_executable_path():
5018 from zipimport
import zipimporter
5019 if hasattr(sys
, 'frozen'): # Running from PyInstaller
5020 path
= os
.path
.dirname(sys
.executable
)
5021 elif isinstance(globals().get('__loader__'), zipimporter
): # Running from ZIP
5022 path
= os
.path
.join(os
.path
.dirname(__file__
), '../..')
5024 path
= os
.path
.join(os
.path
.dirname(__file__
), '..')
5025 return os
.path
.abspath(path
)
5028 def load_plugins(name
, suffix
, namespace
):
5031 plugins_spec
= importlib
.util
.spec_from_file_location(
5032 name
, os
.path
.join(get_executable_path(), 'ytdlp_plugins', name
, '__init__.py'))
5033 plugins
= importlib
.util
.module_from_spec(plugins_spec
)
5034 sys
.modules
[plugins_spec
.name
] = plugins
5035 plugins_spec
.loader
.exec_module(plugins
)
5036 for name
in dir(plugins
):
5037 if name
in namespace
:
5039 if not name
.endswith(suffix
):
5041 klass
= getattr(plugins
, name
)
5042 classes
[name
] = namespace
[name
] = klass
5043 except FileNotFoundError
:
5049 obj
, *path_list
, default
=None, expected_type
=None, get_all
=True,
5050 casesense
=True, is_user_input
=False, traverse_string
=False):
5051 ''' Traverse nested list/dict/tuple
5052 @param path_list A list of paths which are checked one by one.
5053 Each path is a list of keys where each key is a string,
5054 a function, a tuple of strings/None or "...".
5055 When a fuction is given, it takes the key as argument and
5056 returns whether the key matches or not. When a tuple is given,
5057 all the keys given in the tuple are traversed, and
5058 "..." traverses all the keys in the object
5059 "None" returns the object without traversal
5060 @param default Default value to return
5061 @param expected_type Only accept final value of this type (Can also be any callable)
5062 @param get_all Return all the values obtained from a path or only the first one
5063 @param casesense Whether to consider dictionary keys as case sensitive
5064 @param is_user_input Whether the keys are generated from user input. If True,
5065 strings are converted to int/slice if necessary
5066 @param traverse_string Whether to traverse inside strings. If True, any
5067 non-compatible object will also be converted into a string
5071 _lower
= lambda k
: (k
.lower() if isinstance(k
, str) else k
)
5072 path_list
= (map(_lower
, variadic(path
)) for path
in path_list
)
5074 def _traverse_obj(obj
, path
, _current_depth
=0):
5076 path
= tuple(variadic(path
))
5077 for i
, key
in enumerate(path
):
5078 if None in (key
, obj
):
5080 if isinstance(key
, (list, tuple)):
5081 obj
= [_traverse_obj(obj
, sub_key
, _current_depth
) for sub_key
in key
]
5084 obj
= (obj
.values() if isinstance(obj
, dict)
5085 else obj
if isinstance(obj
, (list, tuple, LazyList
))
5086 else str(obj
) if traverse_string
else [])
5088 depth
= max(depth
, _current_depth
)
5089 return [_traverse_obj(inner_obj
, path
[i
+ 1:], _current_depth
) for inner_obj
in obj
]
5091 if isinstance(obj
, (list, tuple, LazyList
)):
5092 obj
= enumerate(obj
)
5093 elif isinstance(obj
, dict):
5096 if not traverse_string
:
5100 depth
= max(depth
, _current_depth
)
5101 return [_traverse_obj(v
, path
[i
+ 1:], _current_depth
) for k
, v
in obj
if key(k
)]
5102 elif isinstance(obj
, dict) and not (is_user_input
and key
== ':'):
5103 obj
= (obj
.get(key
) if casesense
or (key
in obj
)
5104 else next((v
for k
, v
in obj
.items() if _lower(k
) == key
), None))
5107 key
= (int_or_none(key
) if ':' not in key
5108 else slice(*map(int_or_none
, key
.split(':'))))
5109 if key
== slice(None):
5110 return _traverse_obj(obj
, (..., *path
[i
+ 1:]), _current_depth
)
5111 if not isinstance(key
, (int, slice)):
5113 if not isinstance(obj
, (list, tuple, LazyList
)):
5114 if not traverse_string
:
5123 if isinstance(expected_type
, type):
5124 type_test
= lambda val
: val
if isinstance(val
, expected_type
) else None
5125 elif expected_type
is not None:
5126 type_test
= expected_type
5128 type_test
= lambda val
: val
5130 for path
in path_list
:
5132 val
= _traverse_obj(obj
, path
)
5135 for _
in range(depth
- 1):
5136 val
= itertools
.chain
.from_iterable(v
for v
in val
if v
is not None)
5137 val
= [v
for v
in map(type_test
, val
) if v
is not None]
5139 return val
if get_all
else val
[0]
5141 val
= type_test(val
)
5147 def traverse_dict(dictn
, keys
, casesense
=True):
5148 write_string('DeprecationWarning: yt_dlp.utils.traverse_dict is deprecated '
5149 'and may be removed in a future version. Use yt_dlp.utils.traverse_obj instead')
5150 return traverse_obj(dictn
, keys
, casesense
=casesense
, is_user_input
=True, traverse_string
=True)
5153 def variadic(x
, allowed_types
=(str, bytes, dict)):
5154 return x
if isinstance(x
, collections
.abc
.Iterable
) and not isinstance(x
, allowed_types
) else (x
,)
5157 # create a JSON Web Signature (jws) with HS256 algorithm
5158 # the resulting format is in JWS Compact Serialization
5159 # implemented following JWT https://www.rfc-editor.org/rfc/rfc7519.html
5160 # implemented following JWS https://www.rfc-editor.org/rfc/rfc7515.html
5161 def jwt_encode_hs256(payload_data
, key
, headers
={}):
5167 header_data
.update(headers
)
5168 header_b64
= base64
.b64encode(json
.dumps(header_data
).encode('utf-8'))
5169 payload_b64
= base64
.b64encode(json
.dumps(payload_data
).encode('utf-8'))
5170 h
= hmac
.new(key
.encode('utf-8'), header_b64
+ b
'.' + payload_b64
, hashlib
.sha256
)
5171 signature_b64
= base64
.b64encode(h
.digest())
5172 token
= header_b64
+ b
'.' + payload_b64
+ b
'.' + signature_b64
5176 # can be extended in future to verify the signature and parse header and return the algorithm used if it's not HS256
5177 def jwt_decode_hs256(jwt
):
5178 header_b64
, payload_b64
, signature_b64
= jwt
.split('.')
5179 payload_data
= json
.loads(base64
.urlsafe_b64decode(payload_b64
))
5183 def supports_terminal_sequences(stream
):
5184 if compat_os_name
== 'nt':
5185 from .compat
import WINDOWS_VT_MODE
# Must be imported locally
5186 if not WINDOWS_VT_MODE
or get_windows_version() < (10, 0, 10586):
5188 elif not os
.getenv('TERM'):
5191 return stream
.isatty()
5192 except BaseException
:
5196 _terminal_sequences_re
= re
.compile('\033\\[[^m]+m')
5199 def remove_terminal_sequences(string
):
5200 return _terminal_sequences_re
.sub('', string
)
5203 def number_of_digits(number
):
5204 return len('%d' % number
)
5207 def join_nonempty(*values
, delim
='-', from_dict
=None):
5208 if from_dict
is not None:
5209 values
= map(from_dict
.get
, values
)
5210 return delim
.join(map(str, filter(None, values
)))
5216 __initialized
= False
5218 def __init__(self
, parser
, label
=None):
5219 self
._parser
, self
.label
= parser
, label
5220 self
._loaded
_paths
, self
.configs
= set(), []
5222 def init(self
, args
=None, filename
=None):
5223 assert not self
.__initialized
5226 location
= os
.path
.realpath(filename
)
5227 directory
= os
.path
.dirname(location
)
5228 if location
in self
._loaded
_paths
:
5230 self
._loaded
_paths
.add(location
)
5232 self
.__initialized
= True
5233 self
.own_args
, self
.filename
= args
, filename
5234 for location
in self
._parser
.parse_args(args
)[0].config_locations
or []:
5235 location
= os
.path
.join(directory
, expand_path(location
))
5236 if os
.path
.isdir(location
):
5237 location
= os
.path
.join(location
, 'yt-dlp.conf')
5238 if not os
.path
.exists(location
):
5239 self
._parser
.error(f
'config location {location} does not exist')
5240 self
.append_config(self
.read_file(location
), location
)
5244 label
= join_nonempty(
5245 self
.label
, 'config', f
'"{self.filename}"' if self
.filename
else '',
5247 return join_nonempty(
5248 self
.own_args
is not None and f
'{label[0].upper()}{label[1:]}: {self.hide_login_info(self.own_args)}',
5249 *(f
'\n{c}'.replace('\n', '\n| ')[1:] for c
in self
.configs
),
5253 def read_file(filename
, default
=[]):
5255 optionf
= open(filename
)
5257 return default
# silently skip if file is not present
5259 # FIXME: https://github.com/ytdl-org/youtube-dl/commit/dfe5fa49aed02cf36ba9f743b11b0903554b5e56
5260 contents
= optionf
.read()
5261 if sys
.version_info
< (3,):
5262 contents
= contents
.decode(preferredencoding())
5263 res
= compat_shlex_split(contents
, comments
=True)
5269 def hide_login_info(opts
):
5270 PRIVATE_OPTS
= set(['-p', '--password', '-u', '--username', '--video-password', '--ap-password', '--ap-username'])
5271 eqre
= re
.compile('^(?P<key>' + ('|'.join(re
.escape(po
) for po
in PRIVATE_OPTS
)) + ')=.+$')
5276 return m
.group('key') + '=PRIVATE'
5280 opts
= list(map(_scrub_eq
, opts
))
5281 for idx
, opt
in enumerate(opts
):
5282 if opt
in PRIVATE_OPTS
and idx
+ 1 < len(opts
):
5283 opts
[idx
+ 1] = 'PRIVATE'
5286 def append_config(self
, *args
, label
=None):
5287 config
= type(self
)(self
._parser
, label
)
5288 config
._loaded
_paths
= self
._loaded
_paths
5289 if config
.init(*args
):
5290 self
.configs
.append(config
)
5294 for config
in reversed(self
.configs
):
5295 yield from config
.all_args
5296 yield from self
.own_args
or []
5298 def parse_args(self
):
5299 return self
._parser
.parse_args(list(self
.all_args
))